jhnnsjkbk commited on
Commit
5d59930
1 Parent(s): 34fb751

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +53 -0
README.md CHANGED
@@ -1,3 +1,56 @@
1
  ---
2
  license: apache-2.0
 
 
 
 
 
 
 
 
 
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: apache-2.0
3
+ language:
4
+ - en
5
+ tags:
6
+ - Pytorch
7
+ - mmsegmentation
8
+ - segmentation
9
+ - Flood mapping
10
+ - Sentinel-2
11
+ - Geospatial
12
+ - Foundation model
13
+ metrics:
14
+ - accuracy
15
+ - IoU
16
  ---
17
+ ### Model and Inputs
18
+ The pretrained [Prithvi-100m](https://huggingface.co/ibm-nasa-geospatial/Prithvi-100M/blob/main/README.md) parameter model is finetuned to segment the extend of floods on Sentinel-2 images from the [Sen1Floods11 dataset](https://github.com/cloudtostreet/Sen1Floods11).
19
+
20
+ The dataset consists of 446 labeled 512x512 chips that span all 14 biomes, 357 ecoregions, and 6 continents of the world across 11 flood events. The benchmark associated to Sen1Floods11 provides results for fully convolutional neural networks trained in various input/labeled data setups, considering Sentinel-1 and Sentinel-2 imagery.
21
+
22
+ We extract the following bands for flood mapping:
23
+
24
+ 1. Blue
25
+ 2. Green
26
+ 3. Red
27
+ 4. Narrow NIR
28
+ 5. SWIR 1
29
+ 6. SWIR 2
30
+
31
+ Labels represent no water (class 0), water/flood (class 1), and no data/clouds (class 2).
32
+
33
+ The Prithvi-100m model was initially pretrained using a sequence length of 3 timesteps. Based on the characteristics of this benchmark dataset, we focus on single-timestamp segmentation here. This demonstrates that our model does not require multiple timestamps during finetuning.
34
+
35
+
36
+
37
+ ### Code
38
+ Code for Finetuning is available through [github](https://github.com/NASA-IMPACT/hls-foundation-os/)
39
+
40
+ Configuration used for finetuning is available through this [config](https://github.com/NASA-IMPACT/hls-foundation-os/blob/main/fine-tuning-examples/configs/sen1floods11.py).
41
+
42
+ ### Results
43
+ The experiment by running the mmseg stack for 80 epochs using the above config led to the following result:
44
+
45
+ | **Classes** | **IoU**| **Acc**|
46
+ |:------------------:|:------:|:------:|
47
+ | No water | 96.90% | 98.11% |
48
+ | Water/Flood | 80.46% | 90.54% |
49
+
50
+ |**aAcc**|**mIoU**|**mAcc**|
51
+ |:------:|:------:|:------:|
52
+ | 97.25% | 88.68% | 94.37% |
53
+
54
+
55
+ ### Inference
56
+ The github repo includes an inference script that allows to run the flood mapping model for inference on Sentinel-2 images. These input have to be geotiff format, including 6 bands for a single time-step described above (Blue, Green, Red, Narrow NIR, SWIR, SWIR 2) in order. There is also a **demo** that leverages the same code **[here](https://huggingface.co/spaces/ibm-nasa-geospatial/Prithvi-100M-sen1floods11-demo)**.