diff --git "a/data/video_summarization/tvsum/instruct_vhd_50_tvsum_15asr.json" "b/data/video_summarization/tvsum/instruct_vhd_50_tvsum_15asr.json"
new file mode 100644--- /dev/null
+++ "b/data/video_summarization/tvsum/instruct_vhd_50_tvsum_15asr.json"
@@ -0,0 +1 @@
+[{"video": "TVSum/videos/AwmHb44_ouw.mp4", "QA": [{"q": "Transcribed speech: 11.4 - 16.5 seconds, motorcycle, you know, in small vehicle tires just with a couple of screw divers or something. 32.2 - 35.0 seconds, rubber bead of the tire to the outside of the rim. 52.2 - 59.9 seconds, You may need to squish the tire in a vice, press it in a tire machine, or place the. 74.4 - 80.2 seconds, Another method is put the rubber part of the tire on your face of your jack and put it underneath. 102.2 - 108.8 seconds, You step on your tire and you get to bead to drop to the lowest pocket inside your. 127.9 - 133.0 seconds, It gets in the way and sometimes causes you not to be able to get your bead to drop deeply. 150.5 - 155.5 seconds, something like that and I lube the edges like so. 176.1 - 179.5 seconds, unpeeling it off or peeling it back on. 195.1 - 199.4 seconds, If you want to completely take the tire off the rim it's best to just clip it over like. 214.8 - 220.0 seconds, or even putting the old one back in, tubes very often have a right or left side unless. 235.0 - 238.1 seconds, You've got your tire all put back together. 248.8 - 252.0 seconds, Got the valve stem over there, a biscrap opposite to it. 273.4 - 279.0 seconds, Now especially on quad tires to reseal the bead takes a lot of pressure because it's. 303.1 - 304.1 seconds, And it's done. 317.4 - 319.0 seconds, to break the bead. 336.7 - 341.2 seconds, So if you don't have one of those kind of machines you better go to a place to have someone.  Based on the video content and possible transcribed speech, Watch the provided video from the tvsum dataset and mark out the timestamps with stand-out visual content. Document the timestamps of these highlights and evaluate their saliency scores.", "a": "The highlight timestamps are in the 0.0, 1.5, 96.1, 96.6, 99.1, 102.1, 103.1, 106.1, 109.1, 110.1, 120.1, 123.1, 125.6, 128.1, 132.1, 138.1, 138.6, 142.1, 142.6, 143.6, 147.6, 152.2, 152.7, 153.7, 156.2, 159.7, 164.2, 165.7, 172.7, 174.2, 176.2, 188.2, 192.2, 196.2, 198.2, 198.7, 200.2, 222.2, 248.7, 250.2, 256.3, 257.8, 260.8, 274.3, 275.3, 284.3, 284.8, 300.3 seconds. Their saliency scores are 2.9, 2.9, 3.0, 3.0, 2.5, 3.1, 3.1, 2.7, 2.6, 2.6, 3.0, 2.8, 2.7, 2.5, 2.6, 2.6, 2.6, 3.1, 3.1, 3.1, 2.2, 3.0, 3.0, 3.0, 3.8, 3.0, 3.0, 3.0, 2.7, 2.4, 2.6, 3.8, 2.7, 2.5, 3.5, 3.5, 2.6, 2.6, 2.5, 2.5, 2.8, 2.8, 2.7, 3.5, 3.5, 3.0, 3.0, 2.7."}]}, {"video": "TVSum/videos/akI8YFjEmUw.mp4", "QA": [{"q": "Transcribed speech: 12.9 - 17.3 seconds, This means that you get to play a role in reducing greenhouse gas emissions by driving. 19.0 - 24.6 seconds, An exhibition took place to showcase a wide range of electric cars and related equipment. 24.6 - 27.3 seconds, Let's take a look. 27.3 - 32.2 seconds, This is an electric car that emits less carbon dioxide and air pollutants. 32.2 - 37.6 seconds, If you connect this car with a charger, it provides electric current through a transmission. 37.6 - 42.4 seconds, The Minnesota Environment sold metropolitan government and Korea Environment Corporation. 42.4 - 46.7 seconds, have directly established infrastructure for electric cars. 46.7 - 53.1 seconds, Last year we established a charging system to feed up to 500 electric cars. 53.1 - 58.0 seconds, We planned to build another one with a capacity of 2600 vehicles this year and it takes. 58.0 - 65.3 seconds, place more electric cars will come into place for to lead to carbon emissions reduction and air quality improvement. 65.3 - 69.8 seconds, This is a machine that access a state of electric voltage and can't for efficient use. 69.8 - 71.9 seconds, of electric vehicles. 71.9 - 78.2 seconds, As it can estimate the efficiency rate, it's useful when conducting research in development. 78.2 - 84.8 seconds, So 2012, the Electric Vehicle and Smart Car exhibition was held on October 17th to display eco-friendly. 84.8 - 87.6 seconds, electric vehicles and related equipment. 87.6 - 93.8 seconds, Many factors of electric vehicles related industries including two wheel vehicles, parts. 93.8 - 97.7 seconds, and materials, charging systems participated in this exhibition. 97.7 - 101.8 seconds, This is a business fair that is aimed at facilitating exchanges between professionals. 101.8 - 104.6 seconds, in the electric vehicle industry. 104.6 - 109.7 seconds, This fair presented diverse electric vehicles, auto parts and analysis system as well as all. 109.7 - 112.6 seconds, electric battery powered voices. 112.6 - 116.9 seconds, Overses market seminars were held on the sideline of the event to give information to. 116.9 - 120.4 seconds, Korean electric car manufacturers. 120.4 - 124.8 seconds, Electric vehicles had reduced greenhouse gas emissions and air pollutants. 124.8 - 129.5 seconds, To raise penetration rate of electric cars, companies will have to endure to develop.  Based on the video content and possible transcribed speech, You are given a video from the tvsum dataset. Please find the highlight contents in the video, determining the highlight timestamps and its saliency score on a scale from 1 to 5. The output format should be like: 'The highlight timestamps are in the 82, 84, 86, 88, 90, 92, 94, 96, 98, 100 second. Their saliency scores are 1.3, 1.7, 1.7, 1.7, 1.7, 1.3, 1.7, 2.3, 2.3, 2.3'.", "a": "The highlight timestamps are in the 26.5, 32.5, 33.0, 35.0, 38.0, 44.0, 45.5, 66.1, 67.1, 72.1, 76.1, 110.1, 110.6, 122.1, 124.6, 126.1, 128.1, 130.6 seconds. Their saliency scores are 2.4, 2.7, 2.7, 2.7, 2.7, 2.5, 2.5, 2.6, 2.6, 2.3, 2.5, 2.6, 2.6, 3.0, 2.5, 2.4, 3.1, 2.3."}]}, {"video": "TVSum/videos/i3wAGJaaktw.mp4", "QA": [{"q": "Transcribed speech: 7.6 - 11.1 seconds, The cages for insurance are stainless steel with that nary cages. 15.1 - 18.4 seconds, not sitting in, you know, because see it goes... 24.5 - 29.0 seconds, So they tend to lay down rather than sit up with their tables. 32.7 - 38.7 seconds, So here, if you want a demonstration rosy, for instance, if a dog is hurt or heavy, it's. 43.7 - 48.7 seconds, You can hurt the dog's hips, shoulders, and then it's bad for the groomer, you know, no. 53.1 - 61.0 seconds, They're really fun, really good at the state of the yard, they rise. 66.8 - 72.5 seconds, Some bark a lot because he's inside, so we try to minimize any stress, any stress. 75.7 - 81.0 seconds, We just take a moment to come for the animal, calm him down. 88.0 - 89.0 seconds, You know, cheesier. 99.6 - 104.5 seconds, We try to sell really quick quality for green, free, in particular, and there's... 108.8 - 111.6 seconds, It's what dogs have been eating for thousands of years. 114.1 - 120.0 seconds, We look for foods that are, you know, with like, not factory, farm, salmon, but rather wild. 126.3 - 127.3 seconds, issues for dogs. 131.6 - 134.2 seconds, Their bodies can't process, and they become ill. 141.4 - 146.4 seconds, really good quality, few trees, and we welcome feedback.  Based on the video content and possible transcribed speech, Go through the video content from tvsum dataset, and upon identifying highlight moments, list their timestamps. Subsequently, provide a saliency score for each identified highlight.", "a": "The highlight timestamps are in the 5.0, 5.5, 10.0, 28.0, 32.5, 34.0, 43.0, 50.1, 56.1, 64.1, 88.1, 89.6, 91.1, 92.1, 92.6, 93.6, 95.1, 98.1, 111.6, 148.6 seconds. Their saliency scores are 3.2, 3.2, 2.9, 2.8, 2.6, 2.6, 2.5, 2.5, 3.2, 2.8, 3.3, 3.3, 2.9, 3.6, 3.6, 3.6, 3.3, 3.3, 1.2, 1.8."}]}, {"video": "TVSum/videos/Bhxk-O1Y7Ho.mp4", "QA": [{"q": "Transcribed speech: 7.0 - 10.6 seconds, Good morning. Happy Saturday. I do not look my best right now. 35.2 - 40.2 seconds, myself so we can give him a bath. Well now I ordered the wall of bravara. 61.4 - 63.4 seconds, Figure I'll just get this part out of the way. 84.6 - 89.6 seconds, Enjoy the results that I get with that. I think I'm just gonna I'm less. 115.2 - 117.2 seconds, He has like a little. 145.5 - 148.0 seconds, How did he synthesize him and I think it worked a little bit?. 169.5 - 174.4 seconds, I know my momma's going to do it. She my momma's gonna keep me looking fly. 206.1 - 208.1 seconds, There's the concierge. 237.8 - 240.9 seconds, Like a family you understand how parking works. 265.4 - 269.5 seconds, You can hear cut we need to fix that. 307.2 - 309.2 seconds, Look at all that hair. 333.8 - 339.6 seconds, Oh my god, I like totally jacked up his butthole. Do you like it, Bobis? Do we like your haircuts?. 363.6 - 367.7 seconds, His natural oils come back. It'll blend a lot better. He looks so cute. 396.8 - 397.6 seconds, I love this. 419.3 - 423.0 seconds, But I'm just too tired to talk about it. So I'm gonna say bye. 449.0 - 451.0 seconds, Oh, you're such a sweet boy, obviously.  Based on the video content and possible transcribed speech, You are given a video from the tvsum dataset. Please find the highlight contents in the video, determining the highlight timestamps and its saliency score on a scale from 1 to 5. The output format should be like: 'The highlight timestamps are in the 82, 84, 86, 88, 90, 92, 94, 96, 98, 100 second. Their saliency scores are 1.3, 1.7, 1.7, 1.7, 1.7, 1.3, 1.7, 2.3, 2.3, 2.3'.", "a": "The highlight timestamps are in the 0.0, 70.1, 71.6, 74.1, 78.1, 86.6, 110.1, 110.6, 118.1, 121.1, 123.1, 148.1, 149.6, 152.2, 154.2, 162.2, 164.2, 166.2, 168.7, 170.2, 174.2, 200.2, 268.3, 272.3, 274.8, 276.3, 286.3, 294.3, 294.8, 299.3, 326.3, 332.3, 336.8, 338.3, 339.8, 344.3, 396.4, 426.4, 427.4, 434.4, 435.9, 438.4, 438.9, 442.4, 448.4, 449.4 seconds. Their saliency scores are 3.0, 2.7, 2.7, 3.0, 3.2, 2.7, 3.2, 3.2, 3.0, 2.8, 2.4, 2.5, 2.5, 2.5, 2.5, 2.8, 2.9, 2.9, 2.8, 2.5, 3.3, 2.5, 2.8, 3.4, 2.4, 2.8, 2.9, 2.9, 2.9, 2.8, 3.0, 3.0, 3.0, 3.0, 3.0, 2.5, 2.2, 2.5, 2.5, 3.0, 3.0, 3.2, 3.2, 3.2, 2.8, 2.8."}]}, {"video": "TVSum/videos/0tmA_C6XwfM.mp4", "QA": [{"q": "Transcribed speech: 13.0 - 18.8 seconds, However, some animals don't, and this can be a problem. 18.8 - 24.0 seconds, Cutting the nails can be a difficult problem for both of you. 24.0 - 29.9 seconds, Initially, it seems that it will be impossible, but with love and patience it can be done. 29.9 - 34.6 seconds, Sometimes your dog or cat may be uncomfortable when you are handling parts of the body. 34.6 - 37.6 seconds, that he isn't used to having touched. 37.6 - 42.4 seconds, Get him to lie down, try to relax him, and slowly get him used to allowing you to touch. 42.4 - 46.1 seconds, his ears, legs, feet, and paws. 46.1 - 51.4 seconds, This will be helpful for when you need to visit the vet or your groomer. 51.4 - 54.9 seconds, Like humans, dogs have two parts to their nails. 54.9 - 56.7 seconds, The end of the nail is dead. 56.7 - 60.6 seconds, The other part, closest to the paw, has small veins. 60.6 - 65.0 seconds, Sometimes you can see where the vein is, but this depends on the color of the nail. 65.0 - 74.4 seconds, It is important not to cut too close to the paw, as this will make him bleed. 74.4 - 79.4 seconds, The first time you cut your dog's nails, let him smell the scissors or nail clippers. 79.4 - 81.2 seconds, Let him get used to them. 81.3 - 87.8 seconds, Slowly, when the dog is relaxed, cut one nail at a time, talking to him the whole time. 87.8 - 91.4 seconds, Consult your vet or pet shop if you have any questions. 97.4 - 102.4 seconds, Make sure you look for the dew claw, and cut the nail because sometimes this can cause. 102.4 - 107.5 seconds, problems and become embedded in the skin. 107.5 - 111.2 seconds, Glorion nail clippers and scissors remain very sharp. 111.2 - 115.8 seconds, Other clippers or scissors blunt very quickly, and instead of cutting the nails, they can. 115.8 - 117.4 seconds, splinter or break them. 117.4 - 122.7 seconds, If your clippers or scissors do not cut properly, please change them for new ones. 122.7 - 127.3 seconds, Your dog will love that you spend time taking care of him. 127.3 - 129.0 seconds, Do you need more information?. 129.0 - 130.4 seconds, Do not hesitate. 130.4 - 132.0 seconds, Contact us at Gloria. 132.0 - 133.0 seconds, We are here to help.  Based on the video content and possible transcribed speech, Watch the provided video from the tvsum dataset and mark out the timestamps with stand-out visual content. Document the timestamps of these highlights and evaluate their saliency scores.", "a": "The highlight timestamps are in the 28.2, 36.6, 52.2, 56.4, 60.6, 61.2, 64.2, 72.6, 73.8, 76.8, 100.2 seconds. Their saliency scores are 2.9, 2.5, 2.9, 3.0, 3.5, 3.5, 3.2, 3.0, 3.0, 2.8, 2.0."}]}, {"video": "TVSum/videos/3eYKfiOEJNs.mp4", "QA": [{"q": "Transcribed speech: 8.0 - 11.0 seconds, Everybody wants to look good. 11.0 - 14.0 seconds, But dressing up isn't just for portenias. 14.0 - 19.0 seconds, Being out on the town presentable is just as important for the dogs. 19.0 - 24.0 seconds, But if you're a dog, where do you go for your beauty treatin' toots?. 24.0 - 31.0 seconds, At Sebastian's, in Rekuletta, dogs big and small come in for a model makeover. 31.0 - 57.0 seconds, They are hundreds of dog beauty farmers in the city, here at Sebastian's, the land to see East Kugo. 57.0 - 60.0 seconds, There is no service that he won't do. 88.0 - 92.0 seconds, Hugo, rooms and styles around 25 dogs a day. 93.0 - 97.0 seconds, But some customers meet more convincing than others. 118.0 - 121.0 seconds, He is a very good person. 130.0 - 131.0 seconds, He is a good person. 132.0 - 134.0 seconds, Down here, everyone. 135.0 - 137.0 seconds, He is the boss. 138.0 - 141.0 seconds, He is a great player. 143.0 - 144.0 seconds, And another step quoted another two other puppy. 144.0 - 145.0 seconds, He is a great player. 145.0 - 146.0 seconds, And ever he will stop. 161.0 - 170.1 seconds, By me, my entire goal is to successfully continue to destroy his new crew. 171.1 - 172.8 seconds, In the public meeting we also have an appointment Springs. 173.3 - 174.3 seconds, and Amelia ... 176.7 - 178.7 seconds, It is very quirky. 180.5 - 181.3 seconds, We are watching this event.  Based on the video content and possible transcribed speech, From the tvsum dataset, generate a summarized version of the video, focusing on extracting key frames that best represent the overall narrative. The output should be a list of timestamps in seconds and their corresponding salient scores", "a": "The highlight timestamps are in the 50.4, 54.6, 55.8, 60.6, 64.2, 64.8, 66.0, 76.2, 78.0, 82.2, 84.6, 88.2, 96.6, 135.6, 138.6, 170.4, 172.2, 176.4, 178.8 seconds. Their saliency scores are 3.1, 3.2, 3.2, 2.8, 2.8, 2.8, 2.8, 2.8, 2.8, 3.2, 3.5, 3.1, 2.5, 1.9, 2.5, 3.0, 3.2, 3.0, 2.5."}]}, {"video": "TVSum/videos/xxdtq8mxegs.mp4", "QA": [{"q": "Transcribed speech: 26.2 - 29.4 seconds, We're going to talk a little bit today about your health and your cleaning. 29.4 - 34.0 seconds, The dogs develop dirt and debris and discharge in their ears, and it has to be taken out. 34.0 - 38.8 seconds, And so we use solutions like the ear cleansing solution to help break up debris and dirt,. 38.8 - 42.4 seconds, to help change the pH of the ears, to help make it where it's not a hospitable environment. 42.4 - 46.0 seconds, for the bugs, and also to help with the odor. 46.0 - 48.8 seconds, We use ear cleaning solutions in addition to treatments as well. 48.8 - 53.2 seconds, The cleaning solution helps kill bacteria in yeast and can be used afterwards for regular cleaning. 53.2 - 56.7 seconds, One thing to remember is that if you're treating and cleaning the ears at the same time, you need. 56.7 - 58.7 seconds, to clean the ears 30 minutes prior to treatment. 59.4 - 66.0 seconds, So, saltier has been kind enough today to be our model for our ear cleaner. 66.0 - 67.8 seconds, A couple of tips on cleaning ears. 67.8 - 72.6 seconds, First of all, we want to grip the ear at the base of the ear, but not so much that we're causing pain. 72.6 - 75.0 seconds, Second of all, take your ear cleaning solution. 75.0 - 77.8 seconds, Apply a couple of drops into the ear canal. 77.8 - 82.6 seconds, Massage it very well, making sure to work the solution into the ear canal. 82.6 - 88.4 seconds, And then take some cotton swabs and wipe away any debris from the outside. 88.4 - 90.4 seconds, A few more tips for cleaning your dog's ears. 90.4 - 96.0 seconds, We're done with the cleaning, wipe off the tip of the ear cleaning solution to make sure there's no dirt and debris left on there. 96.0 - 100.2 seconds, When applying it to the dog's ear, make sure not to jam the actual tip into the ear. 100.2 - 106.7 seconds, And when cleaning the ears, make sure not to jam any Q-tips or cotton into the actual ear, only wipe off the debris from the outside. 106.7 - 113.7 seconds, If there are multiple pets in the house, it would be a good idea to have multiple bottles so you don't spread micros back and forth from dog to dog. 113.7 - 118.2 seconds, If you have any questions about ear cleaning, your local veterinarian is a great source for information.  Based on the video content and possible transcribed speech, Identify and extract the most emotionally impactful moments from the video provided by tvsum dataset, rating their intensity on a scale from 1 to 5.", "a": "The highlight timestamps are in the 23.0, 34.5, 35.0, 36.5, 45.5, 59.0, 64.5, 66.5, 68.5, 72.5, 73.5, 76.5, 78.5, 82.5, 83.0, 88.0, 88.5, 90.5, 94.5, 97.5 seconds. Their saliency scores are 1.9, 2.3, 2.3, 2.5, 1.9, 2.0, 2.2, 2.4, 2.5, 3.2, 3.2, 3.4, 3.9, 3.6, 3.6, 2.9, 2.4, 2.6, 2.5, 2.5."}]}, {"video": "TVSum/videos/WG0MBPpPC6I.mp4", "QA": [{"q": "Transcribed speech: 3.0 - 9.8 seconds, I am making a chicken torta and I'm using a special bootleg tip bread crumb with kotija cheese and cilantro. 26.3 - 31.0 seconds, You can use well-hoc and cheese or any type of panell cheese that you want and if you can't find that use mozzarella. 47.5 - 50.7 seconds, You can use any type of tomatoes. It's a tomato. It's a tomato. It's not necessarily. 64.4 - 66.4 seconds, I'm gonna let you know. 86.2 - 88.2 seconds, Get this on the plate like that. 101.6 - 104.2 seconds, Bread and that's it and we're gonna go to the fryer. 126.2 - 130.3 seconds, You're not to the plate. We're gonna do just one more piece because we want a very. 150.2 - 157.8 seconds, As always get extra fat into my meals. How am I gonna do that? I'm going to butter my bread and make it. 167.4 - 169.7 seconds, And it goes my bread meanwhile. 189.4 - 199.8 seconds, And just do a nice little heat smash so I have some how you say lightly fried beans. 215.9 - 221.7 seconds, That's right. You can't all right my beans are about done. Give it a little more smashy smash. 245.2 - 249.1 seconds, Okay, we are now going to do something that takes a lot of concentration. 266.3 - 268.3 seconds, There you go tomatoes. 308.0 - 311.2 seconds, And I'm gonna give it a nice little crunch so it kind of breaks it up. 326.9 - 333.3 seconds, We're just gonna lay it across. Okay, I cut my lettuce on I got my jalapeno peppers and that's right. 353.1 - 355.1 seconds, sandwich is a sandwich is. 366.9 - 368.9 seconds, There we go. 387.2 - 390.7 seconds, Hello, I'm chef Diane DeMailan welcome to bootleg tips today.  Based on the video content and possible transcribed speech, Go through the video content from tvsum dataset, and upon identifying highlight moments, list their timestamps. Subsequently, provide a saliency score for each identified highlight.", "a": "The highlight timestamps are in the 8.1, 52.6, 82.6, 83.8, 106.4, 126.4, 168.3, 180.8, 190.2, 192.7, 200.2, 222.1, 230.3, 242.1, 258.4, 264.0, 266.5, 269.7, 276.6, 286.6, 300.3, 324.1, 326.6, 338.5, 339.1, 344.1, 352.3, 354.1, 356.6, 366.7, 367.3, 370.4, 387.3, 394.2 seconds. Their saliency scores are 2.6, 2.9, 2.4, 2.4, 2.6, 2.6, 2.9, 2.5, 2.9, 2.7, 2.4, 2.6, 2.0, 2.8, 2.6, 2.5, 2.8, 2.6, 3.2, 3.0, 3.4, 2.9, 3.3, 3.2, 3.2, 3.0, 2.9, 2.3, 2.5, 2.6, 2.6, 2.5, 1.7, 2.5."}]}, {"video": "TVSum/videos/Hl-__g2gn_A.mp4", "QA": [{"q": "Transcribed speech: 3.7 - 7.6 seconds, Today we are making something super delicious as always. 15.9 - 17.8 seconds, We have some homemade pickles. 27.4 - 33.2 seconds, mustard here because I have to take little tiny bites of the corn beef in between. 45.4 - 46.4 seconds, You don't have to use rye bread. 51.3 - 52.6 seconds, Next, we're going to start building. 59.4 - 61.4 seconds, There we go. 92.1 - 95.6 seconds, Russian dressing, but we're going to do the Russian dressing on the side. 101.9 - 104.7 seconds, All that really is, it's just some ketchup and mayonnaise. 115.6 - 117.6 seconds, Kind of like a thousand-dollar dressing. 134.4 - 139.0 seconds, And then we're going to take it to the stove, get it hot on one side, flip that bad boy. 148.8 - 150.4 seconds, My pan is nice and hot. 162.8 - 164.0 seconds, all over that. 175.2 - 178.0 seconds, Open sandwich out of the pan onto my plate. 191.8 - 196.0 seconds, Now it's time to taste it, but I almost forgot. 203.1 - 204.1 seconds, You are my friend. 221.2 - 223.4 seconds, Booyah, that was a touchdown. 237.6 - 240.4 seconds, Hello, I'm Chef Diane DeMe and welcome to my kitchen.  Based on the video content and possible transcribed speech, Identify and extract the most emotionally impactful moments from the video provided by tvsum dataset, rating their intensity on a scale from 1 to 5.", "a": "The highlight timestamps are in the 0.0, 8.1, 60.1, 72.6, 73.2, 78.2, 82.0, 84.5, 112.6, 114.5, 128.3, 128.9, 150.8, 166.4, 168.3, 170.2, 172.0, 172.7, 176.4, 177.1, 178.3, 187.1, 188.3, 206.5, 226.5, 227.7 seconds. Their saliency scores are 1.3, 3.0, 2.5, 3.8, 3.8, 3.2, 2.4, 3.0, 2.8, 2.2, 2.8, 2.8, 2.0, 3.4, 3.5, 3.7, 3.9, 3.9, 2.4, 2.4, 2.4, 2.7, 2.5, 3.0, 3.3, 3.3."}]}, {"video": "TVSum/videos/Yi4Ij2NM7U4.mp4", "QA": [{"q": "Transcribed speech: 4.1 - 8.2 seconds, War Man's Meals so let's go ahead and get right into it. I was thinking to. 14.8 - 21.2 seconds, Yeah that's not my name. Okay? Um what is your name then?. 28.9 - 33.3 seconds, I've got some hot links. I've got some ketchup. I've got some onions. What do you think?. 38.9 - 45.7 seconds, Obviously myself is being stubborn or turned Ferguson. So bottom one is this. We're. 56.9 - 67.7 seconds, Cheek. 60 cents. 60 cents. Cheek. But what do you need in order to cut those onions and peppers? Sharp stuff. 87.3 - 103.3 seconds, Cut it. And cut it good. Peel that. Mmm. Cut onions. Be in a man. Mmm. 113.9 - 123.9 seconds, Jeez. Next thing you don't get any burn your face off. You break. Don't need a hole in it. 136.7 - 155.1 seconds, Don't want it too thick either. Nice and thin slices. Not like the big stupid bite that I took. Small. Green pepper. Green pepper. Take the label off. Labels are nasty. Not good for one's health. 173.9 - 189.9 seconds, And it wouldn't be spicy sausage sandwiches without. Boom. Spicy hot links. The best. How much would these back? One buck. A buck. 200.9 - 219.9 seconds, Yeah. That's right. Look at me cover the label with my hand. Why? Because they're not a sponsor. That's why. How much was the bottle? Two bucks. The whole meal costs less than five bucks. Now that's a poor man's meal. Let's cook it. If you can't stand the heat, get out of the kitchen. 250.9 - 262.9 seconds, Add links. Write in. That's what we're talking about. Pepper's an onions right on top. Boom. Don't want to catch up to boil over. 292.9 - 314.9 seconds, It cooks for 45 minutes right here on the stove. This kind of thing, you could stick in a crock pot and cook for seven hours. Wouldn't recommend cooking any longer, but still cooking for as long as you need or think necessary. Mmm, it's good. Low heat. Don't want to over boil it, but you can see that catch up. Colosian already. Turns in eating like a natural barbecue sauce with those green peppers and onions. Good stuff. 334.9 - 349.9 seconds, It's like that. I'm there. Just a self spoon, but it's awesome veggies. Mix it up a little bit here. There we go. 368.9 - 377.9 seconds, And bam. Just like that spicy sausage sandwiches. Yum. And freaking hot. Let it cool down a little bit. 392.9 - 398.9 seconds, So I hope you enjoyed it. Join me next time for another episode of Poor Man's Meals. Until then, keep on eating.  Based on the video content and possible transcribed speech, You are given a video from the tvsum dataset. Please find the highlight contents in the video, determining the highlight timestamps and its saliency score on a scale from 1 to 5. The output format should be like: 'The highlight timestamps are in the 82, 84, 86, 88, 90, 92, 94, 96, 98, 100 second. Their saliency scores are 1.3, 1.7, 1.7, 1.7, 1.7, 1.3, 1.7, 2.3, 2.3, 2.3'.", "a": "The highlight timestamps are in the 53.8, 58.8, 61.2, 120.6, 124.4, 125.6, 130.6, 132.5, 134.4, 136.2, 138.1, 142.5, 156.2, 158.1, 160.6, 164.4, 166.2, 168.1, 175.0, 183.1, 236.2, 240.6, 243.1, 245.6, 250.6, 256.2, 260.6, 268.8, 293.1, 298.1, 298.8, 334.4, 368.8, 372.5, 374.4, 375.6, 376.9, 380.0 seconds. Their saliency scores are 2.5, 2.9, 2.6, 2.5, 2.5, 2.5, 2.5, 2.4, 2.5, 2.8, 2.5, 2.5, 2.6, 2.9, 2.5, 2.5, 2.6, 2.8, 2.6, 1.9, 2.6, 2.6, 2.4, 2.4, 2.4, 3.2, 3.4, 3.2, 2.5, 2.9, 2.9, 2.8, 2.5, 3.3, 3.0, 3.0, 2.6, 2.4."}]}, {"video": "TVSum/videos/37rzWOQsNIw.mp4", "QA": [{"q": "Transcribed speech: 7.0 - 24.0 seconds, This is the tofu sandwich. 24.0 - 32.0 seconds, This is very nice. 32.0 - 39.0 seconds, We are just moving around a bit. 39.0 - 42.0 seconds, No, it is. 43.0 - 44.0 seconds, Asian art museum. 44.0 - 51.0 seconds, The English sandwich. 51.0 - 58.0 seconds, How is it?. 66.0 - 69.0 seconds, This is the mixed combination. 69.0 - 72.0 seconds, Go away birds. 72.0 - 77.0 seconds, Birds too close. 99.0 - 108.0 seconds, So we can kind ofusalemheath hijacked a whole band. 108.0 - 110.0 seconds, So kind of deserted this. 110.0 - 112.0 seconds, This is Vietnamese being putting. 113.0 - 115.0 seconds, It's called chair. 134.0 - 135.0 seconds, How is it?. 137.0 - 139.0 seconds, Really good digging. 141.0 - 143.0 seconds, So really this rice inside. 145.0 - 146.0 seconds, What is this?. 147.0 - 149.0 seconds, It's very interesting. 151.0 - 153.0 seconds, What is the meat out of it?. 153.0 - 157.0 seconds, There's rice and it's wrapped inside a banana leaf. 157.0 - 159.0 seconds, It's stuffed pork and steamed. 160.0 - 161.0 seconds, It's really good. 175.0 - 177.0 seconds, How is it?. 183.0 - 186.0 seconds, This is likely a meat version of the evil castle.  Based on the video content and possible transcribed speech, In the video presented from tvsum dataset, seek moments that could serve as an executive summary for a busy stakeholder. It's vital to notate their timestamps and to score each based on their level of saliency.", "a": "The highlight timestamps are in the 0.0, 4.0, 8.0, 8.5, 10.5, 18.0, 18.5, 21.5, 30.0, 38.0, 45.0, 66.1, 70.1, 74.1, 75.6, 152.2, 154.2, 159.2, 162.2, 181.2 seconds. Their saliency scores are 3.2, 2.9, 2.5, 2.5, 2.4, 3.1, 3.1, 3.0, 2.5, 2.4, 2.1, 3.1, 3.0, 2.8, 2.8, 2.4, 2.6, 2.5, 2.5, 2.3."}]}, {"video": "TVSum/videos/98MoyGZKHXc.mp4", "QA": [{"q": "Transcribed speech: 8.3 - 13.6 seconds, And while these kits aren't always popular with drivers, if your car has one, you'll need. 21.6 - 27.1 seconds, If you get a puncture, pull over and stop as soon as you find a safe place, preferably. 29.2 - 36.2 seconds, By the handbrake and switch your engine off, and turn on your hazard lights if necessary. 38.6 - 43.8 seconds, Sealant kits will only fill holes up to about 4mm across, so if you've got a nail in. 45.8 - 50.8 seconds, But if the tires torn or the alloy wheel itself is damaged, you're going to need to call. 54.2 - 59.5 seconds, You'll find your repair kit in the boot, it'll either be in the side panels or as. 61.3 - 67.6 seconds, Now it consists of two parts, the bottle of sealant and the compressor for pumping up. 74.4 - 78.4 seconds, the valve cap and attach the sealant bottle. 83.5 - 88.2 seconds, On this Toyota, the air compressor from flating the tires is then attached to the back. 89.2 - 93.8 seconds, However, on many other cars, you'll need to unscrew the bottle and attach the air compressor. 97.1 - 99.6 seconds, Now it's time to inflate your tire. 104.6 - 106.4 seconds, inside your car. 108.8 - 113.7 seconds, Now turn the compressor on and inflate the tire until it reaches the recommended pressure. 117.3 - 122.3 seconds, If the tire isn't inflating properly, try rolling the car forward a few metres to help. 124.6 - 128.4 seconds, If that doesn't work, you may still need to call a breakdown service. 140.4 - 145.0 seconds, You can carry on driving, but you'll need to stop a couple of miles down the road to. 147.6 - 150.8 seconds, And remember, this is only a temporary fix. 156.0 - 160.1 seconds, When you get to a garage, remember that you'll need to buy a new can of sealant and a new. 166.4 - 172.4 seconds, So check our reviews before you buy. 177.8 - 180.5 seconds, and find out which breakdown companies we recommend.  Based on the video content and possible transcribed speech, Identify and extract the most emotionally impactful moments from the video provided by tvsum dataset, rating their intensity on a scale from 1 to 5.", "a": "The highlight timestamps are in the 28.2, 28.8, 66.6, 69.0, 76.2, 76.8, 82.2, 82.8, 98.4, 104.4, 106.8, 108.6, 116.4, 117.0, 128.4, 130.2 seconds. Their saliency scores are 3.1, 3.1, 4.3, 3.8, 3.0, 3.0, 3.2, 3.2, 2.8, 3.5, 3.3, 2.6, 3.2, 3.2, 2.9, 2.6."}]}, {"video": "TVSum/videos/LRw_obCPUt0.mp4", "QA": [{"q": "Transcribed speech: 19.0 - 24.2 seconds, you have this thing you want to eat and you got to blow on it for like a few seconds and. 30.0 - 35.1 seconds, with your finger and be like all right that's ready but it's trapping heat that. 40.6 - 47.2 seconds, lock the temperature texture ratio dialed in or when you hand it to somebody at a. 52.4 - 57.9 seconds, getting that texture on the outside the moisture on the inside and the heat. 64.3 - 68.7 seconds, the main key factors I look for in a fried chicken sandwich. 74.9 - 79.5 seconds, Village what makes our fried chicken sandwich different it's a great. 86.2 - 91.3 seconds, simplicity you can do it really really well or you can do it half-hazard. 95.6 - 99.7 seconds, the city although there are some other fried chickens out there that I aspire to. 105.3 - 112.1 seconds, season flour and it's crisp it's awesome you know I would go toe-to-toe with. 116.0 - 119.7 seconds, going to bring some acid to my chicken sandwich I want something that's cold. 124.7 - 131.0 seconds, it's going to bring nice acid it's going to balance the temperatures as well. 138.6 - 143.3 seconds, enhance and explode in your mouth. I think toppings make it unique I think just. 148.2 - 153.3 seconds, that the spiciness and the Tabasco you know the crunch of the cucumber and the. 158.4 - 162.6 seconds, it's one of those things we've tried taking it off the menu and we've had a. 169.5 - 183.4 seconds, around. My philosophy on food I think ultimately no matter what it boils down. 189.8 - 196.5 seconds, we're proud of that makes the person eating it smile. This is a damn good sandwich. 203.6 - 208.8 seconds, America, I'm American, part of my family is American and then within that you know. 215.3 - 220.9 seconds, hard to make connections with our guests. We are trying really really hard to. 228.1 - 234.8 seconds, in our door and then if you're doing it well then maybe that's unique.  Based on the video content and possible transcribed speech, From the tvsum dataset, generate a summarized version of the video, focusing on extracting key frames that best represent the overall narrative. The output should be a list of timestamps in seconds and their corresponding salient scores", "a": "The highlight timestamps are in the 2.5, 4.4, 6.3, 8.1, 11.3, 56.3, 58.2, 60.1, 62.6, 65.1, 66.3, 108.2, 118.2, 128.3, 137.6, 150.1, 158.3, 198.3, 198.9, 200.8, 236.5, 240.2 seconds. Their saliency scores are 2.6, 2.9, 2.8, 2.5, 1.6, 2.5, 2.6, 2.7, 3.0, 2.6, 3.0, 2.4, 3.0, 3.6, 3.6, 2.6, 3.5, 2.6, 2.6, 2.3, 2.9, 3.5."}]}, {"video": "TVSum/videos/cjibtmSLxQ4.mp4", "QA": [{"q": "Transcribed speech: 51.4 - 69.9 seconds, do the same for you and she left the. 125.7 - 127.7 seconds, I'm not sure if I'm going to do anything. 222.5 - 230.9 seconds, but they tend to be nice it says and.. 279.5 - 280.7 seconds, This is going to be your load. 317.3 - 318.3 seconds, Every time until this. 333.3 - 336.3 seconds, This is thepoint of his horse. 352.0 - 353.3 seconds, He can't follow. 389.5 - 396.6 seconds, I would entertain young people. 418.9 - 423.1 seconds, to claim one counsel to seem to recognise. 443.8 - 447.7 seconds, and then to her son,. 469.3 - 473.3 seconds, I think it's a very important thing to do with the people. 487.3 - 490.3 seconds, Do with the people who have been in the same place. 515.3 - 520.9 seconds, No it sounds like everyone took it to you. 549.3 - 551.6 seconds, When I did a depression I couldn't bother. 575.3 - 577.3 seconds, wait walking Kai!. 604.9 - 606.3 seconds, This is the two set of fur.  Based on the video content and possible transcribed speech, From the tvsum dataset, generate a summarized version of the video, focusing on extracting key frames that best represent the overall narrative. The output should be a list of timestamps in seconds and their corresponding salient scores", "a": "The highlight timestamps are in the 9.0, 21.0, 30.0, 38.5, 44.0, 44.5, 60.6, 74.1, 82.1, 100.1, 102.1, 103.1, 124.1, 146.1, 147.6, 150.2, 151.7, 154.2, 172.2, 174.2, 176.2, 179.7, 190.2, 191.2, 198.2, 199.2, 200.2, 201.7, 204.7, 232.2, 258.3, 258.8, 259.8, 262.8, 272.3, 274.3, 275.8, 308.3, 314.3, 322.8, 380.4, 386.4, 388.4, 518.0, 519.5, 524.0, 530.0, 536.0, 537.0, 554.1, 555.1, 580.1, 584.1, 588.1, 590.1, 591.1, 606.1, 612.1, 620.1, 620.6, 626.6, 628.1, 631.1 seconds. Their saliency scores are 3.5, 2.4, 2.7, 2.6, 3.3, 3.3, 3.2, 3.4, 3.5, 2.5, 2.5, 2.5, 3.5, 2.8, 2.8, 2.8, 2.8, 2.2, 2.5, 3.0, 3.2, 3.1, 3.0, 3.0, 2.7, 2.7, 2.6, 2.6, 2.5, 2.5, 2.6, 2.6, 2.6, 2.6, 2.6, 2.6, 2.6, 2.4, 2.2, 2.5, 3.2, 3.7, 2.6, 3.5, 3.5, 3.8, 3.4, 3.7, 3.7, 2.4, 2.4, 2.8, 2.5, 3.0, 3.2, 3.2, 2.8, 2.3, 2.7, 2.7, 2.5, 2.6, 2.4."}]}, {"video": "TVSum/videos/b626MiF1ew4.mp4", "QA": [{"q": "Transcribed speech: 12.5 - 18.3 seconds, Like really thorough that. 24.2 - 30.0 seconds, about 70 years ago here at UNC Charlotte and sits then we've grown into a lot of. 32.5 - 33.5 seconds, We progress the live. 42.1 - 47.6 seconds, We train this by running, jumping, climbing, vaulting, balancing, things like that. 58.4 - 63.1 seconds, So it's a process of breaking down obstacles like in your head and figuring out the. 69.8 - 74.4 seconds, So you might be physically capable of doing a new jump or something to move but at the. 83.4 - 87.4 seconds, know, when it's right and when you know that it just goes back. 98.4 - 102.4 seconds, I just want everyone really to kind of do like this on each leg. 109.6 - 113.9 seconds, While we're doing this, let's go and warm up our wrists some. 130.7 - 134.8 seconds, I tore two muscles in my knee because I fell off a building and I didn't roll from my. 142.1 - 145.8 seconds, if you've ever seen people do the dive rolls, that's when we'll be teaching you to. 154.8 - 160.8 seconds, Okay, yeah, we'll have to help you with this one. 169.3 - 172.4 seconds, there to support you when you come back down. 181.3 - 186.3 seconds, So coming through. 200.3 - 205.3 seconds, I'm going to get a jet, quite compact. 211.3 - 218.3 seconds, Alright, who's not the best?. 225.3 - 232.3 seconds, There's some episode avatar, this guy's teasing a visit.  Based on the video content and possible transcribed speech, Go through the video content from tvsum dataset, and upon identifying highlight moments, list their timestamps. Subsequently, provide a saliency score for each identified highlight.", "a": "The highlight timestamps are in the 0.0, 18.8, 28.1, 30.6, 36.2, 58.1, 58.8, 61.2, 62.5, 74.4, 80.6, 112.5, 124.4, 152.5, 182.5, 192.5, 212.5, 218.1, 221.2 seconds. Their saliency scores are 2.5, 2.6, 2.4, 2.2, 1.9, 3.3, 3.3, 2.5, 2.6, 2.6, 2.5, 2.5, 3.0, 2.6, 2.5, 3.0, 2.6, 2.9, 2.6."}]}, {"video": "TVSum/videos/XkqCExn6_Us.mp4", "QA": [{"q": "Transcribed speech: None. Based on the video content and possible transcribed speech, You are given a video from the tvsum dataset. Please find the highlight contents in the video, determining the highlight timestamps and its saliency score on a scale from 1 to 5. The output format should be like: 'The highlight timestamps are in the 82, 84, 86, 88, 90, 92, 94, 96, 98, 100 second. Their saliency scores are 1.3, 1.7, 1.7, 1.7, 1.7, 1.3, 1.7, 2.3, 2.3, 2.3'.", "a": "The highlight timestamps are in the 3.0, 12.0, 22.0, 24.5, 26.0, 41.0, 56.1, 64.6, 76.1, 80.1, 82.1, 86.1, 92.1, 100.1, 101.1, 106.1, 114.1 seconds. Their saliency scores are 2.6, 2.5, 3.0, 2.2, 2.8, 2.4, 2.9, 2.7, 2.8, 3.6, 2.8, 2.6, 3.2, 2.6, 2.6, 2.8, 2.3."}]}, {"video": "TVSum/videos/GsAD1KT1xo8.mp4", "QA": [{"q": "Transcribed speech: 28.0 - 32.0 seconds, If you're searching for a way I will, you gotta take the road. 32.0 - 37.0 seconds, If you don't pick the motion moving, you cannot feel the same. 37.0 - 42.0 seconds, In the heart of the city, or the streets of the veins. 42.0 - 47.0 seconds, Oh, when I'm rocking low, oh, when the river's trying, oh. 47.0 - 50.0 seconds, Can you help me go on?. 51.0 - 55.0 seconds, One more time, we're gonna get up this. 55.0 - 58.0 seconds, We're never gonna run and hide. 58.0 - 63.0 seconds, If I get down and say that I wanna change my mind. 63.0 - 68.0 seconds, Oh, me and a circle, so I can change it one more time. 68.0 - 70.0 seconds, Only in a circle, don't you?. 70.0 - 72.0 seconds, Don't you put me in a circle, don't you call?. 72.0 - 96.0 seconds, One more time, we're gonna get up this. 96.0 - 99.0 seconds, We're never gonna run and hide. 99.0 - 104.0 seconds, If I get down and say that I wanna change my mind. 104.0 - 109.0 seconds, Oh, me and a circle, so I can change it one more time. 109.0 - 112.0 seconds, Only in a circle, don't you call?. 129.0 - 138.0 seconds, I don't know where I'm going, but I know where I've been. 138.0 - 143.0 seconds, And there's never in a nowhere, it's not given it. 143.0 - 145.0 seconds, And when the earth is slipping, japan.  Based on the video content and possible transcribed speech, Go through the video content from tvsum dataset, and upon identifying highlight moments, list their timestamps. Subsequently, provide a saliency score for each identified highlight.", "a": "The highlight timestamps are in the 78.1, 86.1, 88.1, 92.1, 94.6, 97.6, 99.6, 114.1, 115.1, 128.1, 136.1 seconds. Their saliency scores are 3.3, 2.8, 2.9, 3.0, 3.0, 2.4, 2.4, 2.6, 2.6, 2.5, 2.9."}]}, {"video": "TVSum/videos/PJrm840pAUI.mp4", "QA": [{"q": "Transcribed speech: 26.4 - 30.4 seconds, I see a gun. 30.4 - 35.4 seconds, I'll be here again. 35.4 - 40.4 seconds, With you, I'm gonna die. 40.4 - 44.4 seconds, I'm gonna die, I'm gonna die. 52.4 - 56.4 seconds, I'm gonna die. 77.4 - 81.3 seconds, I'm the one for the earth. 81.3 - 86.3 seconds, They told me. 86.3 - 89.3 seconds, Things are soon and no. 89.3 - 93.3 seconds, But it's where I wanna go. 93.3 - 96.3 seconds, I'm definitely not alone. 110.3 - 114.3 seconds, I'm the watch of the rain. 114.3 - 119.3 seconds, This time I'll wait. 120.3 - 123.3 seconds, I'm the watch of the rain. 123.3 - 129.3 seconds, This time I'll wait. 132.3 - 138.3 seconds, I'm the watch of the rain. 143.3 - 148.3 seconds, This time I'll wait. 159.3 - 164.3 seconds, I'll be a renegade. 166.3 - 170.3 seconds, I'm the watch of the rain. 170.3 - 174.3 seconds, I'm definitely not alone. 174.3 - 178.3 seconds, I'm the watch of the rain. 224.3 - 228.3 seconds, This time I'll wait. 229.3 - 233.3 seconds, I'm the watch of the rain. 233.3 - 237.3 seconds, This time I'll wait.  Based on the video content and possible transcribed speech, Identify and extract the most emotionally impactful moments from the video provided by tvsum dataset, rating their intensity on a scale from 1 to 5.", "a": "The highlight timestamps are in the 2.5, 3.1, 10.6, 28.1, 32.5, 42.5, 46.2, 47.5, 50.6, 112.5, 114.4, 132.5, 166.3, 174.4, 192.5, 200.6, 241.2, 246.2, 254.4, 257.5, 258.1, 260.6 seconds. Their saliency scores are 2.5, 2.5, 2.0, 2.2, 2.2, 2.5, 2.6, 2.6, 2.1, 2.1, 2.4, 2.2, 2.2, 3.3, 2.2, 3.3, 1.2, 2.4, 2.5, 2.2, 2.4, 2.8."}]}, {"video": "TVSum/videos/91IHQYk1IQM.mp4", "QA": [{"q": "Transcribed speech: None. Based on the video content and possible transcribed speech, Watch the provided video from the tvsum dataset and mark out the timestamps with stand-out visual content. Document the timestamps of these highlights and evaluate their saliency scores.", "a": "The highlight timestamps are in the 10.0, 10.5, 18.0, 38.0, 42.0, 42.5, 44.5, 62.1, 64.1, 66.1, 83.1 seconds. Their saliency scores are 2.2, 2.2, 2.5, 2.8, 3.5, 3.5, 2.9, 2.5, 3.0, 2.2, 3.2."}]}, {"video": "TVSum/videos/RBCABdttQmI.mp4", "QA": [{"q": "Transcribed speech: 5.0 - 11.4 seconds, It's a great part of Mont. 29.0 - 34.0 seconds, I think it's seeing the whole tradition of all the alums,. 51.0 - 55.0 seconds, And we've only ever seen it in pictures because everybody's an undergraduate squad by now. 88.0 - 92.0 seconds, It's played in the same pattern. 127.6 - 129.5 seconds, There was no one where else to come,. 149.2 - 152.1 seconds, This is the part you got out. 173.2 - 175.9 seconds, And it's great fun to see the seniors. 196.1 - 201.9 seconds, It really doesn't matter whether they're the class of 2013. 221.7 - 225.9 seconds, and we carry it together and we transport it to Mary Line's grave. 241.4 - 243.8 seconds, And it's like the intermingling of so many different cultures. 260.6 - 262.6 seconds, and start to move. 277.2 - 280.8 seconds, When we were seniors, we didn't know what was coming. 298.0 - 300.7 seconds, And it's wonderful looking at those young women. 313.9 - 316.1 seconds, So like for me, it's like my connection. 325.9 - 329.6 seconds, We just seen, I think, 100 years behind us too.  Based on the video content and possible transcribed speech, From the tvsum dataset, generate a summarized version of the video, focusing on extracting key frames that best represent the overall narrative. The output should be a list of timestamps in seconds and their corresponding salient scores", "a": "The highlight timestamps are in the 6.0, 6.5, 32.5, 34.5, 68.6, 84.6, 88.1, 94.1, 124.1, 136.1, 144.1, 156.2, 170.2, 216.2, 222.7, 226.2, 232.2, 266.3, 273.3, 274.3, 286.3, 302.3, 312.3, 312.8, 317.8, 326.3, 336.3, 340.3, 340.8, 346.3, 352.4, 354.4, 356.4 seconds. Their saliency scores are 3.5, 3.5, 3.0, 3.0, 2.3, 2.5, 2.5, 2.4, 2.5, 2.4, 2.4, 2.3, 2.2, 3.3, 3.0, 3.2, 3.1, 2.9, 2.5, 2.5, 2.2, 2.4, 3.0, 3.0, 3.0, 2.5, 2.5, 2.8, 2.8, 2.5, 2.9, 2.4, 2.5."}]}, {"video": "TVSum/videos/z_6gVvQb2d0.mp4", "QA": [{"q": "Transcribed speech: 7.0 - 14.0 seconds, Yeah, it's more comfortable around us. 18.0 - 20.0 seconds, This is a really detailed, right?. 22.0 - 23.0 seconds, Yeah, I get on my left. 25.0 - 26.0 seconds, That should show me. 59.0 - 65.0 seconds, I thought it would have been in there soon enough. 69.0 - 70.0 seconds, Oh, yeah, wait. 77.0 - 80.0 seconds, Thank you very much. 96.0 - 98.0 seconds, There's the most people having. 113.0 - 116.0 seconds, I'll just go over here. 154.8 - 161.8 seconds, Go came side go dancing go. 203.5 - 205.5 seconds, Okay, let's go!. 215.5 - 219.5 seconds, Thank you for bringing the cupboard!. 249.5 - 251.5 seconds, I'm here to get a re-turner. 254.5 - 255.5 seconds, I know, the sun. 259.5 - 261.5 seconds, Just to click click click click. 265.5 - 266.5 seconds, All of them!. 272.5 - 274.5 seconds, That was fun!.  Based on the video content and possible transcribed speech, You are given a video from the tvsum dataset. Please find the highlight contents in the video, determining the highlight timestamps and its saliency score on a scale from 1 to 5. The output format should be like: 'The highlight timestamps are in the 82, 84, 86, 88, 90, 92, 94, 96, 98, 100 second. Their saliency scores are 1.3, 1.7, 1.7, 1.7, 1.7, 1.3, 1.7, 2.3, 2.3, 2.3'.", "a": "The highlight timestamps are in the 132.1, 133.1, 134.6, 138.1, 139.6, 142.1, 143.6, 152.7, 154.2, 156.2, 158.2, 160.2, 169.2, 172.2, 190.2, 194.2, 196.2, 202.2, 205.2, 210.2, 212.2, 214.2, 220.7, 224.2, 224.7, 226.7, 230.2, 232.2, 233.7, 236.2, 238.2, 240.2, 246.2, 247.7 seconds. Their saliency scores are 3.3, 3.3, 3.0, 3.0, 3.0, 2.8, 2.8, 3.1, 2.5, 2.3, 2.4, 3.0, 2.6, 4.2, 2.8, 2.3, 3.4, 3.7, 2.6, 2.4, 3.2, 3.0, 2.4, 3.3, 3.3, 2.9, 2.3, 3.2, 3.2, 2.6, 2.4, 2.6, 3.2, 3.2."}]}, {"video": "TVSum/videos/fWutDQy1nnY.mp4", "QA": [{"q": "Transcribed speech: 15.0 - 18.0 seconds, but say or Paul will come and say Mandarin too. 22.0 - 23.0 seconds, no, not Main Street. 29.0 - 32.0 seconds, We're here for the Chinese New Year for Raid. 37.0 - 39.0 seconds, So there's lots of hot junk that's going all around. 41.0 - 44.0 seconds, And, sir, Paul, I don't know where he went. 52.0 - 54.0 seconds, I just realized it's you. 86.0 - 88.0 seconds, I think it's a no-pated way. 93.0 - 95.0 seconds, I'm trying to get a photo. 101.0 - 103.0 seconds, That's a smile. 106.0 - 108.0 seconds, See you in front of. 115.0 - 117.0 seconds, Mustn't have got no accent. 124.0 - 126.0 seconds, Really, it's Donnea Famous. 133.0 - 135.0 seconds, ters, thank you because I couldn't wait. 396.0 - 400.0 seconds, Just a couple of times. 407.0 - 409.0 seconds, Here's the two. 421.0 - 443.8 seconds, That's where I was sitting. 532.1 - 536.9 seconds, all the cool shapes that they have on this map of.  Based on the video content and possible transcribed speech, You are given a video from the tvsum dataset. Please find the highlight contents in the video, determining the highlight timestamps and its saliency score on a scale from 1 to 5. The output format should be like: 'The highlight timestamps are in the 82, 84, 86, 88, 90, 92, 94, 96, 98, 100 second. Their saliency scores are 1.3, 1.7, 1.7, 1.7, 1.7, 1.3, 1.7, 2.3, 2.3, 2.3'.", "a": "The highlight timestamps are in the 90.2, 108.3, 114.3, 148.4, 150.4, 151.9, 156.4, 190.0, 198.0, 200.1, 202.1, 204.6, 208.1, 211.1, 214.6, 220.1, 236.2, 242.2, 254.7, 260.2, 270.2, 276.3, 282.3, 292.3, 296.3, 311.4, 338.4, 341.9, 344.5, 350.5, 351.5, 356.5, 362.0, 364.0, 370.0, 378.0, 382.1, 383.1, 386.1, 412.6 seconds. Their saliency scores are 2.9, 2.6, 2.7, 2.6, 2.8, 2.8, 3.3, 2.9, 3.0, 3.1, 3.3, 3.0, 3.2, 3.0, 2.9, 2.6, 3.0, 3.3, 3.0, 2.6, 3.0, 3.0, 3.0, 2.6, 2.8, 2.5, 2.9, 2.6, 3.0, 3.0, 3.0, 2.9, 2.6, 2.8, 3.0, 2.6, 2.8, 2.8, 2.9, 2.8."}]}, {"video": "TVSum/videos/J0nA4VgnoCo.mp4", "QA": [{"q": "Transcribed speech: 5.0 - 8.0 seconds, So what are we doing, Kay?. 33.0 - 36.0 seconds, I can say that during November. 65.0 - 69.0 seconds, I think luckily your car is not worth a lot. 97.0 - 100.0 seconds, Oh, the poor young guy. 121.0 - 123.0 seconds, And we're going to have some more good stuff. 147.0 - 148.0 seconds, She probably got a flat tire. 172.0 - 175.0 seconds, Yeah, I'm off with a little bit of a layer. 216.0 - 218.0 seconds, Oh, my legs burn. 243.0 - 245.0 seconds, Do you feel manly?. 297.0 - 298.0 seconds, How's it going?. 350.0 - 352.0 seconds, Let Rebecca help you. 390.0 - 393.0 seconds, It's a scarf. 425.0 - 426.0 seconds, Why is it your life?. 449.0 - 450.0 seconds, We did it. 487.0 - 493.0 seconds, What do you see at him?. 555.1 - 558.1 seconds, Look at this, it's a junior.  Based on the video content and possible transcribed speech, Go through the video content from tvsum dataset, and upon identifying highlight moments, list their timestamps. Subsequently, provide a saliency score for each identified highlight.", "a": "The highlight timestamps are in the 26.3, 27.5, 30.7, 31.9, 94.5, 110.1, 172.7, 176.4, 234.6, 242.1, 242.7, 244.6, 248.4, 268.4, 272.1, 272.8, 277.2, 292.2, 300.3, 305.3, 308.4, 314.1, 315.9, 320.3, 320.9, 322.8, 325.3, 328.5, 330.3, 332.8, 338.5, 339.1, 340.3, 342.8, 346.6, 348.5, 355.4, 358.5, 362.2, 364.1, 369.1, 369.7, 374.1, 382.3, 386.0, 410.4, 411.0, 452.3, 453.0, 456.1, 464.2, 466.1, 472.3 seconds. Their saliency scores are 2.6, 2.6, 2.6, 2.6, 3.4, 2.7, 2.4, 2.1, 2.4, 3.0, 3.0, 2.8, 2.8, 2.7, 3.7, 3.7, 2.6, 3.0, 3.2, 2.9, 2.6, 2.6, 2.6, 3.8, 3.8, 3.5, 3.0, 3.0, 3.2, 2.8, 3.4, 3.4, 3.5, 2.9, 2.8, 2.9, 2.8, 2.3, 2.5, 3.0, 2.9, 2.9, 3.0, 2.4, 2.5, 3.2, 3.2, 2.5, 2.5, 2.5, 2.5, 2.5, 2.5."}]}, {"video": "TVSum/videos/4wU_LUjG5Ic.mp4", "QA": [{"q": "Transcribed speech: 9.2 - 14.5 seconds, South Endon Sea and Cambridge in the UK, Middleburg and the province of Zealand in the Netherlands. 14.5 - 19.4 seconds, Beloining France and Court Chicken Belgium are all working together to promote cycling in the. 20.4 - 26.3 seconds, In summer 2013, these cities held bike parade and invited everyone to come along, dress up. 26.3 - 27.5 seconds, and celebrate cycling. 31.0 - 38.0 seconds, When I'm stapless, we're on south end of the cycle line and we like cycling because it's. 38.0 - 45.0 seconds, a pretty nice, nice new frame and have a really good time to be able to get the environment. 46.0 - 49.0 seconds, It's nice to get out on the bike and meet other people. 61.0 - 65.0 seconds, Do you cycle to work?. 67.0 - 70.0 seconds, Mainly every day you go to work for the bus. 74.0 - 76.0 seconds, Hi, I'm Greta Ticher. 76.0 - 78.0 seconds, I love cycling because it's fun. 78.0 - 79.0 seconds, It keeps me thin. 79.0 - 82.0 seconds, It's tea and it keeps my carbon footprint down. 82.0 - 85.0 seconds, I'm Fran and I like cycling because it's tea. 85.0 - 89.0 seconds, I get there fast and I get to keep it. 90.0 - 105.0 seconds, I really have a race bicycle and some would do a day of the professional side. 105.0 - 108.0 seconds, The other one is for leisure and hobby. 108.0 - 115.0 seconds, For me it's also a matter of moving and also staying out stressed actually and the whole. 115.0 - 117.0 seconds, week is a lot of stress and energy. 117.0 - 124.0 seconds, When the weather is a bit fine like today, I say again and it's exciting. 124.0 - 127.0 seconds, I live in Conteque, Belgium. 127.0 - 135.0 seconds, The bike is the best way to find our way in the city and it helps it inspires me as well.  Based on the video content and possible transcribed speech, In the video presented from tvsum dataset, seek moments that could serve as an executive summary for a busy stakeholder. It's vital to notate their timestamps and to score each based on their level of saliency.", "a": "The highlight timestamps are in the 10.0, 11.3, 16.3, 19.4, 20.6, 22.5, 50.0, 51.3, 52.6, 98.2, 104.5, 126.4, 128.3, 128.9, 138.3, 138.9 seconds. Their saliency scores are 3.0, 3.0, 2.6, 2.5, 2.5, 2.1, 2.5, 2.5, 3.4, 2.4, 3.3, 2.2, 2.4, 2.4, 2.5, 2.5."}]}, {"video": "TVSum/videos/VuWGsYPqAX8.mp4", "QA": [{"q": "Transcribed speech: 57.3 - 87.3 seconds, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh, oh. 87.3 - 92.3 seconds, I'm not gonna go, I'm not gonna go. 132.3 - 135.3 seconds, My death seems I'm in me. 135.3 - 138.3 seconds, When I'm almost there. 138.3 - 142.3 seconds, Seems to see me feeling a close number. 143.3 - 147.3 seconds, My death seems I'm in me. 147.3 - 150.3 seconds, When I'm almost there. 150.3 - 154.3 seconds, Seems to see me feeling a close number. 154.3 - 157.3 seconds, When I'm almost there. 157.3 - 161.3 seconds, Seems to see me feeling a close number. 161.3 - 166.3 seconds, When I'm almost there. 166.3 - 171.3 seconds, I'm not gonna go, I'm not gonna go.  Based on the video content and possible transcribed speech, Go through the video content from tvsum dataset, and upon identifying highlight moments, list their timestamps. Subsequently, provide a saliency score for each identified highlight.", "a": "The highlight timestamps are in the 0.0, 41.4, 62.4, 85.8, 86.4, 92.4, 93.0, 103.2, 105.0, 108.6, 114.6, 118.2, 124.2, 125.4, 130.2, 142.8, 172.2, 178.8, 181.2, 182.4, 195.0, 212.4, 216.0 seconds. Their saliency scores are 2.3, 2.4, 3.5, 2.7, 2.9, 3.0, 3.0, 2.5, 2.2, 2.1, 3.0, 2.5, 3.3, 3.3, 3.2, 2.8, 2.2, 3.0, 2.5, 3.5, 3.5, 3.2, 2.1."}]}, {"video": "TVSum/videos/JKpqYvAdIsw.mp4", "QA": [{"q": "Transcribed speech: 16.0 - 18.8 seconds, I see W. 18.8 - 23.2 seconds, 80, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20,. 23.2 - 53.2 seconds, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20. 53.2 - 71.2 seconds, 20, 20, 20, 20, 20 20, 20, 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 30 20 20 20 20 15 20 20 20 20 20 20 20 20 20 20 20 20 20 finish 20 wand. 71.2 - 81.8 seconds, 20 20 20 20 20 vanishing Begin. 81.8 - 84.6 seconds, He should've died. 95.2 - 98.4 seconds, My brother and sparkling. 99.4 - 102.6 seconds, The first person. 102.7 - 106.5 seconds, is a Seattle bouncing. 111.8 - 118.8 seconds, I'm gonna go to the other side of the world.  Based on the video content and possible transcribed speech, Identify and extract the most emotionally impactful moments from the video provided by tvsum dataset, rating their intensity on a scale from 1 to 5.", "a": "The highlight timestamps are in the 0.0, 4.2, 4.8, 16.2, 18.0, 48.0, 54.0, 58.2, 90.6, 109.2, 112.2, 116.4, 124.2, 128.4, 140.4 seconds. Their saliency scores are 2.4, 2.4, 2.4, 2.5, 2.7, 2.9, 2.9, 3.0, 2.5, 2.6, 2.2, 2.4, 2.6, 3.0, 2.5."}]}, {"video": "TVSum/videos/xmEERLqJ2kU.mp4", "QA": [{"q": "Transcribed speech: 37.0 - 49.2 seconds, I'll keep it short, sweet. I want to talk to you this new real quick about what we. 61.0 - 65.0 seconds, Be it something major that happens to your favorite to you or something by. 72.0 - 76.0 seconds, But we say some of our players are always called, but we ask them to consistently. 85.0 - 91.0 seconds, If you think about the play that you're like, don't get caught up in the things that are behind you. 94.0 - 98.0 seconds, And I think it'll really help you with the things you're protecting the people. 106.0 - 108.0 seconds, And I'm about to... 118.0 - 124.0 seconds, There's a try that survives on catching monkeys out on the force. 139.0 - 143.0 seconds, They've stayed with fear now. 151.0 - 156.0 seconds, So the monkey could take their hand and stick it inside the coconut. 164.0 - 166.0 seconds, And the monkey could go to color in and out. 171.0 - 174.0 seconds, All the broken along with the monkey. 180.0 - 183.0 seconds, All that monkey I knew was what. 194.0 - 196.0 seconds, or our players are playing. 200.0 - 202.0 seconds, There's a bad thing that has to have the long way. 206.0 - 208.0 seconds, Key getting after it. 212.0 - 214.0 seconds, You do that in the rest of your life. 407.0 - 409.0 seconds, Good stuff to come.  Based on the video content and possible transcribed speech, You are given a video from the tvsum dataset. Please find the highlight contents in the video, determining the highlight timestamps and its saliency score on a scale from 1 to 5. The output format should be like: 'The highlight timestamps are in the 82, 84, 86, 88, 90, 92, 94, 96, 98, 100 second. Their saliency scores are 1.3, 1.7, 1.7, 1.7, 1.7, 1.3, 1.7, 2.3, 2.3, 2.3'.", "a": "The highlight timestamps are in the 230.2, 234.2, 248.2, 249.2, 252.3, 257.3, 260.3, 260.8, 264.3, 266.3, 267.3, 270.3, 271.3, 272.8, 278.3, 316.3, 326.3, 329.3, 330.3, 334.3, 334.8, 336.8, 338.3, 339.3, 356.4, 368.4, 374.4, 376.4, 378.4, 382.4, 390.4, 396.4, 402.4, 408.4, 412.4, 412.9, 420.4, 440.4 seconds. Their saliency scores are 2.6, 2.7, 2.7, 2.7, 2.9, 2.8, 2.6, 2.6, 2.5, 2.6, 2.6, 2.7, 2.7, 2.4, 2.9, 2.5, 2.9, 2.8, 3.0, 3.2, 3.2, 3.0, 2.9, 2.9, 2.6, 2.5, 3.0, 3.1, 3.0, 3.4, 3.3, 3.2, 3.3, 2.4, 2.6, 2.6, 2.3, 2.5."}]}, {"video": "TVSum/videos/byxOvuiIJV0.mp4", "QA": [{"q": "Transcribed speech: 9.0 - 12.0 seconds, I see warranty 20,000, 40. 12.0 - 14.0 seconds, Are you ready to roll?. 14.0 - 19.0 seconds, In the wonder, action 1616, dig a pretty nation. 19.0 - 20.0 seconds, I'm going to go away with a fight. 20.0 - 24.0 seconds, I'm going to stay the afraid, think it's your direction. 24.0 - 28.0 seconds, Be in a game, dreamy, copy, ring that jungle. 28.0 - 31.0 seconds, I'm going to be on the right, I'm going to go away. 31.0 - 33.0 seconds, I'm going to go away, I'm going to go away. 98.0 - 100.0 seconds, I'm going to go away. 100.0 - 101.0 seconds, I'm going to go away, I'm going to go away.  Based on the video content and possible transcribed speech, In the video presented from tvsum dataset, seek moments that could serve as an executive summary for a busy stakeholder. It's vital to notate their timestamps and to score each based on their level of saliency.", "a": "The highlight timestamps are in the 12.5, 32.5, 38.2, 40.0, 62.6, 66.3, 71.3, 78.2, 86.3, 98.2, 102.6, 103.9, 104.5, 128.3 seconds. Their saliency scores are 2.5, 2.6, 3.5, 2.4, 2.4, 3.3, 1.6, 2.5, 2.5, 3.1, 2.6, 2.6, 2.9, 3.3."}]}, {"video": "TVSum/videos/_xMr-HKMfVA.mp4", "QA": [{"q": "Transcribed speech: 67.1 - 68.1 seconds, I don't know. 68.1 - 70.1 seconds, I'm not sure if I'm going to do it.  Based on the video content and possible transcribed speech, Identify and extract the most emotionally impactful moments from the video provided by tvsum dataset, rating their intensity on a scale from 1 to 5.", "a": "The highlight timestamps are in the 18.0, 21.0, 48.0, 60.1, 69.1, 72.1, 80.1, 80.6, 82.6, 84.1, 85.6, 88.1, 94.1, 98.1, 118.6, 120.6 seconds. Their saliency scores are 2.6, 2.5, 2.2, 2.6, 2.5, 2.5, 3.5, 3.5, 2.6, 3.7, 3.7, 3.0, 3.0, 2.7, 2.6, 2.6."}]}, {"video": "TVSum/videos/WxtbjNsCQ8A.mp4", "QA": [{"q": "Transcribed speech: 10.3 - 21.2 seconds, My daddy was a beekeeper and I kind of followed along with him and after I got to teaching. 34.2 - 37.2 seconds, They go through all those different stages in the high. 47.2 - 52.2 seconds, cans when they are sitting inside down by the cloth here. 59.2 - 67.2 seconds, Now you are going to get in trouble with those bees by opening up the hive. 75.2 - 81.2 seconds, This is what we start with and they can see where they are building the new home. 87.2 - 89.2 seconds, They like my smell. 94.2 - 95.2 seconds, Which one is he?. 99.2 - 100.2 seconds, The size of him. 109.2 - 114.2 seconds, There will be about 60 pounds of honey in that top super. 123.2 - 128.2 seconds, That depends on the honey flow and how many bees you got in the hive. 137.2 - 140.2 seconds, honey flow fill one of these honey supers a week. 145.2 - 148.2 seconds, Approximately 60,000 in August. 157.2 - 158.2 seconds, Is it like a bee?. 164.2 - 169.2 seconds, They are into the sun and the magnetism of the earth. 176.2 - 179.2 seconds, There would be too many bees in the hive. 192.2 - 202.2 seconds, You look, you see some pollen and that's why they tell people to buy honey from a local beekeeper if you have allergies. 211.2 - 213.2 seconds, What is the queen costume?. 219.2 - 223.2 seconds, I just shipped six of these jars down at home, California. 229.2 - 233.2 seconds, I'm able to only mail me the check and the freight.  Based on the video content and possible transcribed speech, In the video presented from tvsum dataset, seek moments that could serve as an executive summary for a busy stakeholder. It's vital to notate their timestamps and to score each based on their level of saliency.", "a": "The highlight timestamps are in the 58.6, 66.1, 66.6, 67.6, 70.1, 72.1, 74.1, 75.6, 77.6, 79.1, 82.1, 83.1, 88.1, 89.6, 92.1, 93.6, 96.1, 97.6, 98.6, 100.1, 118.1, 119.1, 122.1, 124.1, 140.1, 172.2, 173.2, 192.2, 194.2, 222.2, 232.2, 233.7, 238.2, 241.2 seconds. Their saliency scores are 2.1, 3.2, 3.2, 3.2, 2.8, 2.9, 3.0, 3.0, 2.4, 2.4, 3.0, 3.0, 2.4, 2.4, 2.6, 2.6, 3.0, 3.0, 2.6, 2.4, 2.3, 2.3, 2.5, 3.0, 2.7, 3.1, 3.1, 2.3, 2.4, 2.5, 2.9, 2.9, 3.0, 2.4."}]}, {"video": "TVSum/videos/uGu_10sucQo.mp4", "QA": [{"q": "Transcribed speech: None. Based on the video content and possible transcribed speech, You are given a video from the tvsum dataset. Please find the highlight contents in the video, determining the highlight timestamps and its saliency score on a scale from 1 to 5. The output format should be like: 'The highlight timestamps are in the 82, 84, 86, 88, 90, 92, 94, 96, 98, 100 second. Their saliency scores are 1.3, 1.7, 1.7, 1.7, 1.7, 1.3, 1.7, 2.3, 2.3, 2.3'.", "a": "The highlight timestamps are in the 31.2, 36.2, 40.6, 64.4, 76.2, 90.6, 98.1, 100.6, 106.2, 107.5, 112.5, 113.8, 118.1, 124.4 seconds. Their saliency scores are 2.4, 2.8, 2.5, 2.5, 2.5, 2.4, 2.6, 3.0, 3.3, 3.3, 2.6, 2.6, 2.5, 2.4."}]}, {"video": "TVSum/videos/EE-bNr36nyA.mp4", "QA": [{"q": "Transcribed speech: 12.2 - 16.2 seconds, And as a home gardener, you'll probably run into these things at one point in time. 16.2 - 21.7 seconds, Now, this is a little bit of a problem because paper wasps are actually a beneficial. 22.7 - 26.7 seconds, They're out collecting caterpillars off of your crops. 26.7 - 31.9 seconds, What the problem is is when you get too close to them, they're indiscriminate stingers. 31.9 - 35.0 seconds, And you're spring cleaning and getting ready for gardening. 35.0 - 40.5 seconds, It's a good opportunity to take those and just remove them from where they're at. 40.5 - 46.0 seconds, And usually they're found in the eaves of homes and structures and it can be knocked down. 46.0 - 51.7 seconds, with a shovel or other implement and then taken away and thrown away. 51.7 - 56.6 seconds, So if the wasps have built a nest in a location that could potentially be a threat. 56.6 - 61.6 seconds, to yourself or your family or your pets, it's a good idea to go ahead and remove it whenever. 61.6 - 63.2 seconds, you possibly can. 63.2 - 65.0 seconds, And that can be done chemically. 65.0 - 69.7 seconds, There are products available that are specific to wasps and hornet control. 69.7 - 76.0 seconds, And wait until evening, spray that product so that all the wasps are on the nest. 76.0 - 79.5 seconds, And then once they've all died, you can clean that wasp nest off just like you would. 79.5 - 82.6 seconds, this one that's in the eve that we have already removed. 82.6 - 86.0 seconds, This is Jeff Edwards, University of Wyoming Extension from the ground up.  Based on the video content and possible transcribed speech, Go through the video content from tvsum dataset, and upon identifying highlight moments, list their timestamps. Subsequently, provide a saliency score for each identified highlight.", "a": "The highlight timestamps are in the 6.0, 10.5, 20.0, 32.0, 43.5, 46.0, 49.0, 68.1, 80.1, 86.1, 90.1, 94.1 seconds. Their saliency scores are 2.1, 1.5, 1.1, 3.1, 2.7, 2.8, 2.6, 2.5, 2.9, 2.7, 2.4, 2.5."}]}, {"video": "TVSum/videos/Se3oxnaPsz0.mp4", "QA": [{"q": "Transcribed speech: 5.1 - 8.6 seconds, on Kalden new 13 and only on Kalden new 13. 8.6 - 13.8 seconds, Sunuwasu is in Bizbee tonight with people that witnessed this terrible incident. 15.0 - 20.6 seconds, We are inside an outbuilding on a farm here in Bizbee where this massive killer bee hive was. 20.6 - 25.8 seconds, located right up there in the ceiling of this building. You can see they've cut out a hole up there. 25.8 - 33.6 seconds, Right there was a 200 pound hive big enough we're told to hold more than 250,000 killer bees. 34.3 - 39.1 seconds, It is the attack of the killer bees. Sound like a horror flick? This is very real. 39.8 - 43.0 seconds, An animal farm located off Highway 80 in Bizbee. 43.6 - 47.8 seconds, Freelance videographer Al-Signori shot this unbelievable footage,. 47.8 - 51.0 seconds, a thousand pound bore under attack. This is the worst one I've seen. 51.0 - 54.5 seconds, And it got worse when the swarm turned its fury on the farmers. 55.1 - 59.8 seconds, Well I jumped into my truck and unfortunately the passenger window was partly open. 60.2 - 66.1 seconds, So they came in in a black cloud. Many many so I had to swap them out the driver's window. 66.1 - 71.0 seconds, and it was pretty pretty bad and pretty bad. You get some? Four or five times. 71.0 - 75.8 seconds, This hand that did the swatting really looked like I'd been boxing without gloves. 75.8 - 81.6 seconds, When everyone else runs this guy comes buzzing in. Re-booth calls himself the killer bee guy. 81.6 - 86.1 seconds, I do most of the Africanized bee removal for a sheriff police fire border patrol. 86.1 - 91.3 seconds, Booth says the bees seem to be feistier this year. This year is worse than I've seen in 10 years. 91.3 - 96.2 seconds, A thousand pound peg is a huge thing and has skin like that and fat like that on it. Some kind of. 96.2 - 101.0 seconds, surprise that it did kill it. Even in a bee suit and helmet attacking a hive filled with a quarter. 101.0 - 105.8 seconds, million aggressive bees is not easy. And it was extremely hot to get up in there and work because. 105.8 - 110.2 seconds, it was probably 180 degrees. It was it was really something else and I ended up with heatstroke actually. 110.2 - 117.2 seconds, Heatstroke and burning stings. Sadly the 1,000 pound board did not survive. Another pregnant 800 pound. 117.2 - 123.7 seconds, sow went into a coma lost off her babies and is now recovering. Now experts say if you do happen to see. 123.7 - 129.4 seconds, such a hive on your property do not try to eradicate these bees yourself. It is a very dangerous. 129.4 - 134.1 seconds, job that should be left up to professionals. In Bizby Sonu Vasu Kaleldi,. 134.1 - 138.7 seconds, the new 13 live local late breaking. Thank you Sonu. Pima community colleges plan to.  Based on the video content and possible transcribed speech, In the video presented from tvsum dataset, seek moments that could serve as an executive summary for a busy stakeholder. It's vital to notate their timestamps and to score each based on their level of saliency.", "a": "The highlight timestamps are in the 39.5, 42.5, 43.5, 46.5, 48.5, 50.5, 52.5, 74.5, 76.5, 86.5, 87.5, 89.5, 98.5, 102.5, 103.5, 106.5, 112.5, 113.5 seconds. Their saliency scores are 1.9, 3.2, 3.2, 2.5, 2.6, 2.5, 3.3, 2.8, 3.0, 3.5, 3.5, 3.3, 2.6, 3.5, 3.5, 2.8, 3.4, 3.4."}]}, {"video": "TVSum/videos/gzDbaEs1Rlg.mp4", "QA": [{"q": "Transcribed speech: 9.4 - 13.5 seconds, In this short film, I'm going to explain how to repair a tire using the compressor and. 25.7 - 28.2 seconds, allowing the tire to be reinflated. 38.1 - 43.6 seconds, Next, ensure the gear lever is in the neutral position for both manual and automatic transmission. 57.9 - 61.4 seconds, an under-inflated or flat tire for an extended period. 74.3 - 79.1 seconds, spare tire valve and warning sticker in the bag. 91.9 - 97.0 seconds, it pierces the foil. 114.7 - 119.4 seconds, Now we get the fluid and attach the filler tube to the tire valve or the sealant bottle. 131.2 - 136.6 seconds, Disconnect the filler tube from the tire valve and refit the valve insert back into the valve. 154.6 - 160.1 seconds, the inflation tube that will attach to the actual tire itself. 178.3 - 180.6 seconds, The tire will start to inflate. 190.5 - 194.4 seconds, If the pressure stays below the required level, try the car a few metres to spread the. 207.7 - 210.4 seconds, back on and drive straight off. 224.0 - 227.6 seconds, If the tire drops below the pressure is outlined in the owner handbook then it could be. 234.3 - 239.0 seconds, The sealant means that you can drive at a maximum speed of 50mph. 248.9 - 251.0 seconds, replace the damaged tire. 262.1 - 266.8 seconds, And of course if you have any further questions please do not hesitate to contact your.  Based on the video content and possible transcribed speech, Identify and extract the most emotionally impactful moments from the video provided by tvsum dataset, rating their intensity on a scale from 1 to 5.", "a": "The highlight timestamps are in the 34.2, 40.2, 63.0, 71.4, 92.4, 96.6, 98.4, 104.4, 106.2, 106.8, 109.8, 114.6, 120.6, 123.0, 130.8, 136.2, 138.6, 144.6, 148.2, 150.0, 158.4, 159.6, 163.2, 164.4, 172.2, 175.2, 176.4, 186.6, 204.6, 205.2 seconds. Their saliency scores are 2.7, 1.9, 1.3, 2.7, 2.4, 2.4, 2.5, 2.6, 3.0, 3.0, 2.8, 2.7, 2.5, 2.5, 2.5, 2.6, 2.8, 2.9, 3.2, 3.2, 3.5, 3.5, 2.5, 2.5, 3.0, 2.6, 2.2, 2.5, 2.6, 2.6."}]}, {"video": "TVSum/videos/oDXZc0tZe04.mp4", "QA": [{"q": "Transcribed speech: None. Based on the video content and possible transcribed speech, From the tvsum dataset, generate a summarized version of the video, focusing on extracting key frames that best represent the overall narrative. The output should be a list of timestamps in seconds and their corresponding salient scores", "a": "The highlight timestamps are in the 2.5, 92.1, 95.1, 112.1, 112.6, 115.6, 152.2, 152.7, 154.2, 156.2, 206.2, 250.2, 254.3, 262.3, 264.3, 266.3, 285.8, 286.3, 294.3, 296.3, 304.3, 305.8, 308.8, 311.3, 334.3, 338.3, 344.3, 345.3, 348.3, 364.9, 368.9, 370.4 seconds. Their saliency scores are 2.9, 2.5, 2.4, 3.0, 3.0, 2.5, 2.5, 2.5, 2.4, 2.4, 3.3, 2.5, 2.6, 2.2, 2.4, 2.3, 2.2, 2.4, 3.0, 2.6, 3.0, 3.0, 2.6, 2.5, 2.9, 2.6, 2.2, 2.2, 2.2, 2.5, 2.6, 2.1."}]}, {"video": "TVSum/videos/qqR6AEXwxoQ.mp4", "QA": [{"q": "Transcribed speech: 7.4 - 12.2 seconds, Basically whipping the bike is exactly kind of what it sounds like off the face of the jump. 20.7 - 23.6 seconds, Beyond which some riders are able to do on on. 35.6 - 43.2 seconds, But when you do the whip correctly the bike in the momentum of your bike will actually bring you back around and straighten you out before you hit the ground. 57.3 - 59.3 seconds, The key is. 68.7 - 73.7 seconds, Which will bring the bike actually flat as you leave the as your tires leave the. 81.8 - 85.4 seconds, So there's a point where if you actually still framed it where you would be. 97.8 - 104.8 seconds, It'll set your your bike back up straight as long as you do it right the key is just very very small steps. 113.0 - 116.9 seconds, But the landing or the downside where you're shooting for would be at a an angle. 126.4 - 129.0 seconds, That gives you a little bit more room for a margin of error. 139.7 - 141.7 seconds, so we're gonna do a couple and. 184.5 - 189.4 seconds, Taking off with the correct amount of lean and then using your body positioning. 197.7 - 200.6 seconds, The key is just to stay within your ability level. 206.9 - 213.1 seconds, When you're leaning against the jump your front and especially and your rear your rear tire is not gonna break loose on the face. 222.3 - 228.1 seconds, Use your your body and your bike and let the momentum carry it across and you'll be whipping a no time. 240.1 - 244.4 seconds, Analyze, you know the face of the jump find the right jump that you feel comfortable on. 256.3 - 262.1 seconds, Repetitively you'll be able to go a little bit farther a little bit farther and it can stay completely in control.  Based on the video content and possible transcribed speech, In the video presented from tvsum dataset, seek moments that could serve as an executive summary for a busy stakeholder. It's vital to notate their timestamps and to score each based on their level of saliency.", "a": "The highlight timestamps are in the 56.1, 57.1, 58.6, 59.6, 60.1, 74.6, 80.1, 82.1, 83.1, 83.6, 85.1, 92.1, 94.1, 98.1, 119.1, 122.1, 148.1, 149.6, 150.7, 152.2, 152.7, 153.2, 153.7, 154.2, 155.2, 156.2, 157.2, 157.7, 158.2, 159.2, 160.2, 161.2, 162.2, 163.7, 166.2, 166.7, 167.2, 168.2, 168.7, 169.7, 170.7, 172.7, 173.7, 174.7, 176.2, 177.2, 240.2, 240.7, 244.2, 245.2 seconds. Their saliency scores are 4.0, 4.0, 2.4, 2.4, 2.7, 2.2, 2.5, 4.5, 4.5, 4.5, 2.5, 2.1, 2.2, 2.1, 2.5, 2.2, 3.5, 3.5, 2.9, 3.1, 3.1, 3.1, 3.1, 3.4, 3.4, 3.9, 3.9, 3.9, 3.0, 3.0, 2.9, 2.9, 2.3, 2.3, 2.6, 2.6, 2.6, 4.2, 4.2, 4.2, 2.5, 2.5, 2.5, 2.1, 2.2, 2.2, 2.5, 2.5, 2.5, 2.5."}]}, {"video": "TVSum/videos/EYqVtI9YWJA.mp4", "QA": [{"q": "Transcribed speech: 10.7 - 13.3 seconds, I ran into Felipe Smajay. 19.2 - 25.5 seconds, We've been riding since we were around 10 or 11 so I'm about 13 or 14 years. 38.0 - 39.0 seconds, It is a lot of fun. 46.4 - 48.9 seconds, He's the four time national champion. 58.1 - 63.9 seconds, I bet all the like the, like Shaco and gas gas the guys that make Charles bikes probably. 72.6 - 76.0 seconds, thing that we really shoot in Florida because we love that sport and we love to spread it. 88.0 - 92.0 seconds, also the sport of trials and then even more as a whole because it's got a very negative. 101.4 - 105.8 seconds, in a building in downtown LA so like by jumping these barriers we were able to open some. 115.6 - 116.6 seconds, was really good. 128.0 - 131.0 seconds, Who is a while getting here we had some transmission issues. 140.0 - 142.0 seconds, So this is what that feels like. 148.0 - 151.3 seconds, We're going to do the whole series and we've really been able to get our name out there. 163.3 - 164.8 seconds, We're pushing to what you love for a living. 171.7 - 172.7 seconds, We need somebody to jump over. 176.7 - 177.7 seconds, It's had thank you for watching. 180.7 - 186.7 seconds, Get some of these awesome goats shirts we've got.  Based on the video content and possible transcribed speech, In the video presented from tvsum dataset, seek moments that could serve as an executive summary for a busy stakeholder. It's vital to notate their timestamps and to score each based on their level of saliency.", "a": "The highlight timestamps are in the 20.0, 21.5, 24.0, 34.5, 35.5, 46.0, 98.1, 99.6, 114.1, 132.6, 134.1, 134.6, 138.1, 144.1, 144.6, 160.2, 161.7, 165.2, 170.2, 170.7, 176.2, 177.2, 179.2, 180.7, 188.2 seconds. Their saliency scores are 2.7, 2.7, 2.7, 2.7, 2.7, 2.6, 3.3, 3.3, 3.0, 2.6, 2.7, 2.7, 2.8, 3.5, 3.5, 3.0, 3.0, 2.9, 3.0, 3.0, 3.3, 3.3, 3.2, 3.0, 2.8."}]}, {"video": "TVSum/videos/eQu1rNs0an0.mp4", "QA": [{"q": "Transcribed speech: 24.2 - 28.2 seconds, How to stop on a fixed gear bike is the most frequently asked question from. 28.2 - 32.1 seconds, moms, pedestrians or other people who can't believe you're riding a bike with no brakes. 32.1 - 36.5 seconds, There are lots of techniques to stop a fixed gear bike. The one we recommend is. 36.5 - 42.1 seconds, by using your front brake. But if you're going to be a real G and you got rid of. 42.1 - 47.4 seconds, the front brake, here's some tips that will help you slow your roll. First, the. 47.4 - 52.2 seconds, skip stop. This is a technique where you skip on your rear wheel, gradually slowing. 52.2 - 56.7 seconds, you down. When you land, the bike skips and reduces speed. Put a few of these. 56.7 - 60.7 seconds, together and you just save yourself from crashing into that parked car. 60.7 - 65.7 seconds, Hey, what's up guys? I'm New York and if you're wondering how to do a skip stop, I got your answer. 65.7 - 69.4 seconds, First of all, you want to get up to speed. You then want to rise out of your. 69.4 - 72.9 seconds, saddle, which will take weight off the back wheel. Then using your feet, you're. 72.9 - 75.7 seconds, literally going to pick up the back wheel, which will basically stop it in mid-air. 75.7 - 79.6 seconds, Also, keep in mind, whichever foot is forward is the action which you pull up with and. 79.6 - 83.2 seconds, whichever foot is back is the one you push it down with. Basically, those are all the steps. 83.2 - 90.9 seconds, to do as a skip stop. A skit stop is where you lock up your rear wheel, causing your. 90.9 - 97.2 seconds, bike to skid and slow down. This is a fast way to stop, but it also eats up your rear tire. 97.2 - 102.9 seconds, In order to skid, you need to be going somewhat fast. Rise out of your saddle to take. 102.9 - 106.4 seconds, weight off the back wheel. Pull up with your left pedal while pushing down with the right. 106.4 - 110.9 seconds, pedal, keeping in mind that your pedal should be horizontal. If you need extra leverage,. 110.9 - 113.9 seconds, you can always push your leg against the top tube of the frame. That's pretty much how. 113.9 - 123.4 seconds, you do a skid. The easiest way to slow down is to apply back pressure to your pedals. 123.4 - 127.6 seconds, It really looks like nothing, but pull back on the pedals as they rotate and you'll slowly. 127.6 - 132.7 seconds, crease to a stop. When ready fixed gear, it is very important to have some kind of foot. 132.7 - 137.8 seconds, retention on your pedals. You can either use straps or clips, either way you need some kind. 137.8 - 140.3 seconds, foot retention to perform any of the styles we've shown today.  Based on the video content and possible transcribed speech, Watch the provided video from the tvsum dataset and mark out the timestamps with stand-out visual content. Document the timestamps of these highlights and evaluate their saliency scores.", "a": "The highlight timestamps are in the 4.5, 6.0, 7.5, 12.0, 14.0, 17.0, 22.0, 48.0, 50.1, 52.6, 65.6, 67.6, 68.6, 85.1, 100.1, 106.1, 109.1, 110.6, 136.1, 138.1, 139.1 seconds. Their saliency scores are 2.8, 2.1, 2.1, 3.3, 1.9, 2.5, 2.8, 2.7, 3.0, 2.4, 3.0, 2.9, 2.9, 2.0, 2.5, 2.8, 2.8, 2.7, 2.9, 3.2, 3.2."}]}, {"video": "TVSum/videos/JgHubY5Vw3Y.mp4", "QA": [{"q": "Transcribed speech: 6.0 - 16.0 seconds, You never know your power now, but you need to know how to ride. 16.0 - 21.0 seconds, You never know your power now. 25.0 - 31.0 seconds, I'm here with Zach from Pure Fix. Can you give us some expert tips for keeping our ride safe, Zach?. 31.0 - 36.0 seconds, I can, thanks Amanda. A bike is an important investment and you want to keep it safe. 36.0 - 41.0 seconds, The first thing I recommend is getting rid of any quick release levers on your bike. 43.0 - 48.0 seconds, Quick release levers on your wheels and seat post make removing them easy and that makes stealing them easy. 48.0 - 52.0 seconds, It only takes a few seconds to remove a quick release front wheeler saddle. 52.0 - 56.0 seconds, You should replace any quick release levers with Allen bolts or 15mm nuts. 56.0 - 62.0 seconds, This doesn't mean your wheels and saddles are safe, it just makes them harder for a passerby to jack them. 68.0 - 73.0 seconds, If you have a daily commuter that's locked up outside a lot, you might want to consider some camo for your ride. 73.0 - 76.0 seconds, A bike that looks fresh and clean attracts wood-beethaves. 76.0 - 81.0 seconds, Try adding some creative touches to your frame. You can use stickers of your favorite brands to cover up logos,. 81.0 - 84.0 seconds, that'll make it look used even though it may be brand new. 84.0 - 89.0 seconds, Your bike won't ride any different but it'll be much less attractive to thieves. 91.0 - 93.0 seconds, Don't ever leave your bike unlocked. 93.0 - 99.0 seconds, Now to lock your bike using a U-lock, it's the best to attach the lock through your rear wheel around part of the frame. 99.0 - 103.0 seconds, Make sure to lock it to a solid object that can't be broken or moved. 103.0 - 106.0 seconds, And make sure to lock it to your frame, some people forget that. 106.0 - 111.0 seconds, If you're using a chain lock too, loop this through the front wheel and around part of the frame. 111.0 - 114.0 seconds, No lock is ever 100% safe. 114.0 - 119.0 seconds, If someone has enough time and determination, they can cut through just about anything. 119.0 - 123.0 seconds, But if you follow these tips, you can reduce your chances of being another statistic. 123.0 - 129.0 seconds, All right, thank you so much Zach. You guys got to remember to get the right lock and lock up your bike. 129.0 - 133.0 seconds, If you have any comments, please comment below or hit us up on our Twitter. 133.0 - 135.0 seconds, I'm Amanda and I'll see you in the streets. 141.0 - 143.0 seconds, Thank you for watching.  Based on the video content and possible transcribed speech, From the tvsum dataset, generate a summarized version of the video, focusing on extracting key frames that best represent the overall narrative. The output should be a list of timestamps in seconds and their corresponding salient scores", "a": "The highlight timestamps are in the 0.0, 12.5, 54.1, 55.1, 57.1, 72.1, 80.1, 96.1, 100.1, 102.1, 107.6 seconds. Their saliency scores are 1.9, 3.0, 2.4, 2.4, 2.3, 2.4, 2.6, 3.5, 3.5, 3.8, 3.7."}]}, {"video": "TVSum/videos/iVt07TCkFM0.mp4", "QA": [{"q": "Transcribed speech: 5.6 - 8.2 seconds, Today we're gonna teach you how to wheelie. 8.2 - 32.6 seconds, Hey guys, doing a wheelie is something we've all decided to do since our very first bike ride. 32.6 - 34.2 seconds, It's all about balance. 34.2 - 36.8 seconds, Once you get it you can ride the wheelie for miles. 36.8 - 39.0 seconds, My buddy Cole is gonna show us how it's done. 39.0 - 41.2 seconds, Go ahead and do it man. 41.2 - 44.0 seconds, First get comfortable pulling your front wheel up. 44.0 - 46.7 seconds, Then push with your lead foot and lean back. 46.7 - 49.0 seconds, Practice this riding around an empty parking lot. 49.0 - 51.8 seconds, But don't use your straps just in case you have to bail. 51.8 - 53.4 seconds, I want you to able to get the front wheel up. 53.4 - 55.2 seconds, You need to find a balance point. 55.2 - 59.2 seconds, So try maybe going to a wall and holding on and finding that balance point. 59.2 - 60.8 seconds, until you get comfortable. 60.8 - 62.3 seconds, Let's put it all together. 62.3 - 64.6 seconds, Ride forward and pull the front wheel up. 64.6 - 68.6 seconds, Once you push up your lead foot, lean back and find your balance point. 68.6 - 74.1 seconds, Once you are locked in, remember it's a combination of shifting your body weight and the speed that you are pedaling. 82.1 - 83.7 seconds, Thank you George. 83.7 - 86.6 seconds, I'll have to practice that as soon as I'm done here. 86.6 - 90.9 seconds, As always, hit us up on our Twitter if you want to talk or drop a comment below. 90.9 - 93.6 seconds, Until next time, we'll see you in the streets.  Based on the video content and possible transcribed speech, From the tvsum dataset, generate a summarized version of the video, focusing on extracting key frames that best represent the overall narrative. The output should be a list of timestamps in seconds and their corresponding salient scores", "a": "The highlight timestamps are in the 8.8, 11.3, 14.4, 52.6, 58.8, 66.3, 68.2, 72.6, 76.3 seconds. Their saliency scores are 3.0, 2.5, 2.7, 3.2, 2.9, 2.9, 3.3, 3.2, 2.8."}]}, {"video": "TVSum/videos/E11zDS9XGzg.mp4", "QA": [{"q": "Transcribed speech: 2.8 - 5.9 seconds, Look at that, it's on the dohangout.com. 23.3 - 25.2 seconds, It's delicate dormitory. 59.6 - 62.4 seconds, So like, this is in GK-1. 94.6 - 96.6 seconds, Man, any tips you would like to give to your,. 131.6 - 133.6 seconds, Yeah, I'm like a person favorite. 161.6 - 163.6 seconds, according to the agility, the personality,. 189.6 - 190.6 seconds, But, and I'm like,. 221.6 - 224.6 seconds, And you can check out your apparel. 245.6 - 246.6 seconds, I guess so. 268.6 - 269.6 seconds, You're going to buy them. 341.7 - 343.7 seconds, Oh my gosh, I'm there, I'm there, I'm there. 369.7 - 371.7 seconds, We're starting to get off and I'm in the third actually. 399.7 - 401.7 seconds, Where is the GXD. 425.7 - 427.7 seconds, I'm going to go to the same. 461.7 - 463.7 seconds, I'm going to go to the same. 499.7 - 501.7 seconds, Let the peace fall upon us.  Based on the video content and possible transcribed speech, From the tvsum dataset, generate a summarized version of the video, focusing on extracting key frames that best represent the overall narrative. The output should be a list of timestamps in seconds and their corresponding salient scores", "a": "The highlight timestamps are in the 40.5, 43.0, 90.5, 92.5, 112.5, 113.0, 119.5, 124.0, 128.5, 130.5, 133.0, 148.5, 158.5, 162.5, 163.5, 166.0, 170.5, 174.5, 178.5, 248.5, 251.0, 253.0, 256.5, 258.0, 260.5, 262.5, 264.0, 266.0, 268.5, 270.5, 274.5, 288.5, 290.5, 304.5, 306.5, 307.0, 312.5, 318.5, 320.5, 322.5, 325.5, 378.5, 388.5, 391.5, 395.5, 422.5, 428.5, 430.0, 444.5, 445.0, 446.5, 456.5 seconds. Their saliency scores are 3.2, 2.4, 2.5, 2.6, 3.3, 3.3, 2.1, 2.2, 2.4, 2.5, 2.4, 2.3, 2.8, 3.0, 3.0, 2.5, 2.6, 2.5, 2.4, 3.3, 3.0, 2.3, 2.4, 2.4, 2.5, 2.9, 2.9, 2.5, 2.5, 2.8, 2.2, 2.5, 2.5, 2.6, 2.7, 2.7, 2.2, 2.8, 3.1, 2.5, 2.3, 2.5, 2.8, 2.6, 2.2, 2.4, 2.3, 2.3, 2.5, 2.5, 2.7, 2.5."}]}, {"video": "TVSum/videos/NyBmCxDoHJU.mp4", "QA": [{"q": "Transcribed speech: 11.3 - 15.3 seconds, O, o, o. 23.4 - 26.2 seconds, hanging the dark and hadn't had it on the streets fog. 46.7 - 53.7 seconds, What you're I going, that's coming at me through the tree. 64.7 - 66.7 seconds, I don't know what's going to be. 74.7 - 76.7 seconds, Yeah, take me mine. 80.7 - 82.7 seconds, I'm going to make a life. 86.7 - 88.7 seconds, I'm going to shoot the baby. 108.7 - 112.7 seconds, And I don't know what's going to come with me. 118.7 - 120.7 seconds, Let me go. 128.7 - 132.7 seconds, You check my shoes off. 138.7 - 140.7 seconds, I'm going to shoot the baby. 146.7 - 148.7 seconds, You're getting no noise. 152.7 - 154.7 seconds, I'm going to stop the baby. 158.7 - 160.7 seconds, I'm going to let you go. 164.7 - 166.7 seconds, I don't know what's going on. 180.7 - 182.7 seconds, I need to pull the baby.  Based on the video content and possible transcribed speech, Go through the video content from tvsum dataset, and upon identifying highlight moments, list their timestamps. Subsequently, provide a saliency score for each identified highlight.", "a": "The highlight timestamps are in the 13.2, 20.4, 42.6, 43.2, 60.6, 66.6, 72.6, 106.2, 114.6, 115.8, 118.2, 118.8, 136.2, 138.6, 176.4 seconds. Their saliency scores are 2.3, 3.1, 3.0, 3.0, 2.6, 2.9, 2.5, 2.8, 2.9, 2.9, 3.0, 3.0, 2.6, 2.9, 2.8."}]}, {"video": "TVSum/videos/kLxoNp-UchI.mp4", "QA": [{"q": "Transcribed speech: 25.1 - 27.1 seconds, You don't like. 28.1 - 34.1 seconds, And people always trying to hug you. 39.1 - 41.1 seconds, So what do I do?. 45.1 - 49.1 seconds, You can always ignore those people and hide. 50.1 - 53.1 seconds, For as long as you can. 56.1 - 59.1 seconds, And when the door is open just a lot. 61.1 - 63.1 seconds, And take a moment. 70.1 - 72.1 seconds, Explore the ground. 73.1 - 75.1 seconds, Clean up your mess. 76.1 - 78.1 seconds, And just chill. 83.1 - 85.1 seconds, You can take someone. 86.1 - 91.1 seconds, It would not hurt you if you just sit down and think. 96.1 - 98.1 seconds, Before you go inside. 98.1 - 100.1 seconds, And get humiliated. 103.1 - 110.1 seconds, And you can always find a place inside. 110.1 - 113.1 seconds, And enjoy the affection of the giants. 117.1 - 120.1 seconds, It's never too late for living. 121.1 - 124.1 seconds, A wonderful life.  Based on the video content and possible transcribed speech, In the video presented from tvsum dataset, seek moments that could serve as an executive summary for a busy stakeholder. It's vital to notate their timestamps and to score each based on their level of saliency.", "a": "The highlight timestamps are in the 12.0, 22.5, 26.0, 50.1, 58.1, 84.1, 96.1, 101.1, 106.1, 118.1 seconds. Their saliency scores are 3.0, 2.5, 2.7, 2.7, 2.0, 2.7, 2.6, 2.7, 3.0, 2.7."}]}, {"video": "TVSum/videos/jcoYJXDG9sw.mp4", "QA": [{"q": "Transcribed speech: 4.3 - 9.7 seconds, Doxen weighed a whopping 77 pounds. That's twice the normal size, but thanks to. 13.4 - 17.2 seconds, Norah vanada is OB's adopted mom. She's here along with Dr. Andrew Kaplan,. 21.4 - 27.0 seconds, Only you guys. So when you heard about Toby and with OB and you just. 33.2 - 37.3 seconds, match what you were expecting? No. I mean, he was bigger. I had no idea. 43.0 - 46.9 seconds, I hadn't seen a picture and I thought, well, he maybe were expecting a. 51.5 - 55.4 seconds, And I saw him and he really is a purebred Doxen. 59.7 - 63.8 seconds, So Dr. Kaplan, I assume this is the biggest Doxen you've ever seen. 66.2 - 72.3 seconds, I mean, is this, is he healthy? I mean, what are the dangers of a dog weigh in this much?. 77.5 - 80.6 seconds, like atherosclerosis and heart attacks and hypertension,. 85.3 - 94.1 seconds, so a Doxen is, most people know, are predisposed to slip discs and orthopedic problems. 95.8 - 98.6 seconds, Just the disaster. Yeah, I mean, this kind of a weight is going to. 103.8 - 109.0 seconds, What kind of diet is you? We are just putting them on a high protein, high fiber,. 115.5 - 122.1 seconds, And how can I exercise? His only exercise right now is going from house to outside to go potty. 126.2 - 128.9 seconds, And he's moving better every day. Is that a good plan, Doc?. 135.3 - 138.9 seconds, And the fiber helps burn calories and it helps reduce appetite as well. 147.0 - 150.8 seconds, are overweight. Is this an epidemic? We're seeing not only is it in the human population,. 155.4 - 158.9 seconds, feed their dogs. I think what they think they're supposed to feed them, but it's the treats. 163.0 - 168.2 seconds, Is there a correlation between overweight pet owners and overweight pets?. 171.5 - 175.8 seconds, yes, I see a lot of dogs were overweight and their owners tend to be overweight as well. 179.9 - 185.9 seconds, It could take about a year. Our goal is 1% weight loss a week. So far, he's lost about two pounds a week,. 191.6 - 196.0 seconds, Yeah, you should. Absolutely. You're very happy. Are you happy you had told me?.  Based on the video content and possible transcribed speech, In the video presented from tvsum dataset, seek moments that could serve as an executive summary for a busy stakeholder. It's vital to notate their timestamps and to score each based on their level of saliency.", "a": "The highlight timestamps are in the 28.0, 41.0, 51.1, 54.1, 82.1, 89.1, 90.1, 92.1, 93.6, 94.6, 96.1, 97.1, 98.6, 99.6, 102.1, 102.6, 108.1, 108.6, 122.6, 125.6, 128.1, 129.6, 159.2, 164.2, 182.2 seconds. Their saliency scores are 2.4, 2.1, 1.9, 2.4, 2.1, 2.7, 3.5, 3.0, 3.0, 2.7, 3.8, 3.8, 3.2, 3.2, 3.3, 3.3, 2.8, 2.8, 2.9, 2.9, 2.9, 2.9, 3.0, 3.0, 2.3."}]}, {"video": "TVSum/videos/XzYM3PfTM4w.mp4", "QA": [{"q": "Transcribed speech: 10.0 - 11.8 seconds, One that'll continue for a long time. 16.1 - 21.1 seconds, That's why your GMC certified service experts are there to help when it's time for new ones. 25.6 - 28.1 seconds, you can replace all four at once. 32.1 - 35.1 seconds, See how this tire is worn more on one side than the other?. 38.5 - 44.4 seconds, GMC recommends rotating your tires every time you change your oil or 7,500 miles. 49.0 - 51.0 seconds, your tires need replacing. 54.0 - 56.2 seconds, Insert the penny into the tire tread. 59.5 - 64.4 seconds, If not, you want to visit the GMC certified service experts for new tires. 68.4 - 70.4 seconds, You also need proper tire balance. 74.7 - 76.8 seconds, as create uneven tread wear. 81.2 - 82.2 seconds, to lose air. 85.1 - 88.1 seconds, I hope you learned a few things about tire care and replacement. 92.3 - 95.9 seconds, your certified service experts are here to help you select new ones that are right. 97.5 - 99.0 seconds, Don't have your own expert?. 104.2 - 105.6 seconds, Thanks and enjoy your GMC.  Based on the video content and possible transcribed speech, From the tvsum dataset, generate a summarized version of the video, focusing on extracting key frames that best represent the overall narrative. The output should be a list of timestamps in seconds and their corresponding salient scores", "a": "The highlight timestamps are in the 18.5, 20.0, 22.0, 32.5, 42.0, 43.5, 45.0, 50.1, 51.1, 52.1, 56.1, 68.6 seconds. Their saliency scores are 2.6, 2.9, 2.9, 2.9, 2.7, 2.7, 2.6, 2.8, 2.8, 3.2, 3.2, 2.8."}]}, {"video": "TVSum/videos/_esJrBWj2d8.mp4", "QA": [{"q": "Transcribed speech: 12.0 - 18.0 seconds, No? What else do we have? We got some, we got some other kind of food, we got some dog food. 28.0 - 32.0 seconds, You want to try eating some dog food?. 38.0 - 40.0 seconds, Huh? What is this?. 50.0 - 54.0 seconds, I don't believe it, she's eating it. 59.0 - 63.0 seconds, Oh, it's moving around too much on her, so I'm going to get a plate for her. 84.0 - 91.0 seconds, Mevie, Mevie, Mevie, Mevie. 103.0 - 104.0 seconds, Yeah, that's good. 112.0 - 118.0 seconds, So this is kind of strange to see her eating dog food, but she really likes this dog food, so I might have to buy this again. 125.0 - 128.0 seconds, Right! It's good stuff!. 138.0 - 142.0 seconds, Now she's looking at it. 150.0 - 153.0 seconds, We just fed our cat dog food. 157.0 - 160.0 seconds, She did. I knew it. 164.0 - 165.0 seconds, Yeah, it was talking to her, though. 169.0 - 170.0 seconds, Yeah, sure is. 184.0 - 186.0 seconds, She ate dog food and didn't get sick. 190.0 - 193.0 seconds, Taking some good naps today?. 199.0 - 201.0 seconds, Still got your fat belly?. 207.0 - 209.0 seconds, But that's pretty good. 215.0 - 225.0 seconds, So yes, cats can eat dog food.  Based on the video content and possible transcribed speech, You are given a video from the tvsum dataset. Please find the highlight contents in the video, determining the highlight timestamps and its saliency score on a scale from 1 to 5. The output format should be like: 'The highlight timestamps are in the 82, 84, 86, 88, 90, 92, 94, 96, 98, 100 second. Their saliency scores are 1.3, 1.7, 1.7, 1.7, 1.7, 1.3, 1.7, 2.3, 2.3, 2.3'.", "a": "The highlight timestamps are in the 4.5, 18.5, 27.0, 36.5, 44.5, 48.5, 51.0, 56.5, 58.5, 66.5, 68.5, 75.5, 112.5, 115.5, 116.5, 120.0, 122.5, 126.5, 127.0, 206.5, 215.0 seconds. Their saliency scores are 2.4, 2.5, 2.0, 2.2, 3.0, 3.0, 2.7, 2.6, 2.3, 2.9, 3.3, 3.1, 3.0, 2.4, 2.7, 2.4, 2.1, 2.6, 2.6, 2.2, 2.4."}]}, {"video": "TVSum/videos/HT5vyqe0Xaw.mp4", "QA": [{"q": "Transcribed speech: 4.4 - 6.4 seconds, Table which he is. 14.1 - 17.6 seconds, This table just got delivered, but and this is where the story. 32.5 - 37.9 seconds, As you can see the truck ran off the road nearly tipped over went down the gully. 47.2 - 49.2 seconds, Which will surely hold. 54.6 - 56.6 seconds, Let's look down the side shall we?. 64.1 - 67.8 seconds, Here's the dental floss that's holding the truck up sort of. 78.8 - 81.5 seconds, Put so many people out. Look at that. 92.5 - 95.1 seconds, Because there's a truck coming that's gonna try to pull us out. 102.2 - 107.2 seconds, It's great forces aboard to bear on the stuck truck that might made. 114.6 - 120.3 seconds, It stops the driver who's a large husky mary guy led out a ghoulish squeal. 134.6 - 138.1 seconds, We look at a debacle that only mark you and could be responsible for. 149.2 - 151.2 seconds, Because of a table. 164.8 - 166.8 seconds, Low-sarring close as they get. 187.2 - 189.2 seconds, This is the man responsible. 208.2 - 213.3 seconds, As you can see Mark is Mark is basically marinating and stress over this. 288.0 - 290.0 seconds, Here it is. 298.6 - 300.6 seconds, As Mark leaves the scene of the crime. 314.2 - 316.9 seconds, Peel off and with a lot of falling branch.  Based on the video content and possible transcribed speech, From the tvsum dataset, generate a summarized version of the video, focusing on extracting key frames that best represent the overall narrative. The output should be a list of timestamps in seconds and their corresponding salient scores", "a": "The highlight timestamps are in the 39.0, 44.0, 54.6, 58.1, 64.1, 65.1, 66.1, 67.1, 82.1, 84.1, 88.6, 90.1, 110.1, 148.1, 152.2, 160.2, 161.7, 163.2, 164.2, 165.2, 168.2, 169.7, 258.3, 276.3, 280.3, 284.3, 300.3 seconds. Their saliency scores are 2.5, 2.7, 2.3, 2.6, 3.0, 3.0, 3.0, 3.0, 2.8, 2.9, 2.3, 2.0, 2.5, 3.0, 2.2, 4.0, 4.0, 2.4, 2.9, 2.9, 4.0, 4.0, 2.4, 2.5, 2.4, 2.9, 2.6."}]}, {"video": "TVSum/videos/sTEELN-vY30.mp4", "QA": [{"q": "Transcribed speech: 2.4 - 4.6 seconds, We first told you during our four o'clock newscast. 10.5 - 13.2 seconds, It all happened this afternoon where highway 92. 18.3 - 20.0 seconds, from the train crossing tonight. 25.9 - 28.3 seconds, And the driver of that vehicle is thankfully able. 33.1 - 36.2 seconds, Now the problem may be the way that the asphalt was laid. 45.8 - 48.2 seconds, It was just like a boom, you know, screeching. 54.9 - 56.7 seconds, to this wreck of an afternoon. 62.4 - 63.6 seconds, And but that's what happens. 67.9 - 69.8 seconds, you can see what's left of this car. 75.2 - 76.8 seconds, before finally stopping. 83.6 - 86.9 seconds, We've been having truckers get stuck out here,. 95.2 - 98.0 seconds, Broad Street and Mosley Road is Douglasville's most. 103.0 - 106.2 seconds, there have been 22 documented incidents of cars. 110.2 - 113.6 seconds, but left a considerable drop off instead of a tapering slope. 119.2 - 121.3 seconds, No matter how much work they do,. 126.5 - 129.9 seconds, Beverly would rather see the crossing shut down before someone. 132.6 - 136.2 seconds, If I was to see somebody get killed on probably by myself. 142.8 - 145.7 seconds, at this crossing to try and minimize collisions.  Based on the video content and possible transcribed speech, You are given a video from the tvsum dataset. Please find the highlight contents in the video, determining the highlight timestamps and its saliency score on a scale from 1 to 5. The output format should be like: 'The highlight timestamps are in the 82, 84, 86, 88, 90, 92, 94, 96, 98, 100 second. Their saliency scores are 1.3, 1.7, 1.7, 1.7, 1.7, 1.3, 1.7, 2.3, 2.3, 2.3'.", "a": "The highlight timestamps are in the 0.0, 0.5, 10.5, 11.5, 13.5, 45.0, 66.5, 68.5, 70.5, 100.5, 120.5, 130.5, 132.0 seconds. Their saliency scores are 2.9, 2.9, 4.3, 4.3, 4.0, 2.6, 2.8, 3.0, 3.5, 2.2, 2.9, 3.2, 3.2."}]}, {"video": "TVSum/videos/vdmoEJ5YbrQ.mp4", "QA": [{"q": "Transcribed speech: 4.0 - 11.0 seconds, Okay, well, second time driving and we were just like driving down the patch. 29.0 - 34.0 seconds, He was hoping he was going to run out of gas. 47.0 - 54.0 seconds, Oh, that's old snow. 79.0 - 84.0 seconds, We probably stuck for good. 102.0 - 109.0 seconds, That's, yeah, it's, it's pretty bad. 122.0 - 125.0 seconds, It's, it's, no, we can't get it out. 136.0 - 141.0 seconds, But the way things are looking and just take the key out, we'll have to take the Subaru back. 148.0 - 150.0 seconds, Was it hell of an adventure, wasn't it?. 158.0 - 160.0 seconds, Her second day driving. 177.0 - 180.0 seconds, Got an his mouth and I think it did him. 187.0 - 191.0 seconds, No, there's things that go along the edge of the front window. 203.0 - 204.0 seconds, Let's go home. 218.0 - 224.0 seconds, Here comes the cornfield. 234.0 - 237.0 seconds, Here you can go. 248.0 - 249.0 seconds, Oh, it's cold. 272.0 - 273.0 seconds, Where do you go?. 309.0 - 313.0 seconds, See, when you get to be a good driver, you can do this all you want, then you can go to. 320.0 - 323.0 seconds, It's here, so it doesn't help to give it butterflies.  Based on the video content and possible transcribed speech, Go through the video content from tvsum dataset, and upon identifying highlight moments, list their timestamps. Subsequently, provide a saliency score for each identified highlight.", "a": "The highlight timestamps are in the 0.0, 32.0, 76.1, 77.1, 80.1, 84.1, 87.6, 93.6, 104.1, 106.1, 107.1, 109.1, 110.6, 112.1, 115.1, 123.1, 126.1, 127.6, 132.1, 133.1, 136.6, 140.1, 143.6, 144.6, 148.1, 150.2, 150.7, 153.7, 156.7, 168.2, 178.2, 184.2, 186.2, 190.2, 204.7 seconds. Their saliency scores are 2.2, 2.6, 2.1, 2.1, 2.0, 2.7, 2.2, 2.1, 2.9, 2.9, 2.9, 2.7, 2.5, 2.2, 2.0, 2.0, 2.3, 2.3, 3.2, 3.2, 2.4, 3.2, 3.2, 2.5, 2.5, 3.1, 3.1, 3.0, 2.0, 2.2, 2.4, 2.4, 2.5, 2.0, 2.0."}]}, {"video": "TVSum/videos/xwqBXPGE9pQ.mp4", "QA": [{"q": "Transcribed speech: 7.5 - 9.2 seconds, of a car. 18.8 - 20.6 seconds, of riding a motorcycle. 29.0 - 35.5 seconds, You steer it like a car, you sit in it like a car, you have pedals like a car, but the. 50.2 - 52.2 seconds, And this is actually a real gyro right here. 59.9 - 63.2 seconds, So you basically need a baby elf into knock it over. 74.5 - 77.9 seconds, We've just kind of taken that and made it commercially available. 82.1 - 87.0 seconds, But sensors like an IMU, IR sensors. 94.4 - 96.2 seconds, That has to be processed. 105.5 - 111.6 seconds, It's very dependent based on our processing system. 118.7 - 124.7 seconds, What the hurdles in our development is having the vehicle actually be more aware of its surroundings. 138.8 - 143.2 seconds, to allow for stability in all scenarios. 154.6 - 157.5 seconds, And also, it's basically a rolling smartphone. 166.8 - 170.7 seconds, The gyro is actually come in the battery pack coming into its own modular unit. 176.4 - 183.5 seconds, So if you want a faster, more robust, more stabilized, you know, C1, it's pretty easy to. 191.6 - 195.4 seconds, You know, parts count where one tenth of a part of a car. 201.5 - 204.5 seconds, We only have 2,230. 210.2 - 214.7 seconds, In our audience, you know, urban professionals, tech junkies, you know, they have to have the. 223.2 - 224.3 seconds, at affordable price.  Based on the video content and possible transcribed speech, In the video presented from tvsum dataset, seek moments that could serve as an executive summary for a busy stakeholder. It's vital to notate their timestamps and to score each based on their level of saliency.", "a": "The highlight timestamps are in the 20.5, 30.0, 30.5, 33.5, 58.6, 59.6, 72.1, 73.1, 76.1, 76.6, 80.1, 84.1, 85.6, 87.6, 106.6, 120.1, 124.1, 124.6, 125.1, 128.1, 130.1, 161.7, 166.2, 170.2, 174.2, 209.2, 210.2, 224.2, 225.7 seconds. Their saliency scores are 3.2, 3.3, 3.3, 3.2, 2.9, 2.9, 3.0, 3.0, 2.5, 2.5, 2.8, 2.6, 2.6, 2.5, 2.8, 2.5, 3.7, 3.7, 3.7, 3.6, 2.0, 2.4, 2.6, 2.5, 2.5, 2.4, 2.2, 2.5, 2.5."}]}]
\ No newline at end of file