Summary

README.md

@@ -1,8 +1,8 @@
 ---
 title: MASAI
-emoji: 🔝
-colorFrom: gray
-colorTo: red
+emoji: 😀😲😐😥🥴😱😡
+colorFrom: blue
+colorTo: pink
 sdk: gradio
 python_version: 3.12.7
 sdk_version: 5.4.0

app/description.py

@@ -14,4 +14,6 @@ DESCRIPTION = f"""\
 <div class="app-flex-container">
     <img src="https://img.shields.io/badge/version-v{config_data.AppSettings_APP_VERSION}-stable" alt="Version">
 </div>
+
+> MASAI is an advanced tool that assesses and interprets human affective states by integrating data from multiple modalities such as audio, video, text. MASAI accepts multimedia file, extracts audio, video, and text (using ASR), analyzes and combines three modalities, and predicts six basic emotions (Happiness, Sadness, Anger, Surprise, Disgust, Fear), neutral state (Neutral), as well as sentiment (Negative, Positive, Neutral).
 """

app/event_handlers/clear.py

@@ -39,6 +39,6 @@ def event_handler_clear() -> tuple[
         gr.Plot(value=None, visible=False),
         gr.Plot(value=None, visible=False),
         gr.Row(visible=False),
-        gr.Textbox(value=None, visible=False),
-        gr.Textbox(value=None, visible=False),
+        gr.Textbox(value=None, info=None, container=False, visible=False),
+        gr.Textbox(value=None, info=None, container=False, visible=False),
     )

app/event_handlers/event_handlers.py

@@ -30,7 +30,18 @@ def setup_app_event_handlers(
         triggers=[video.change, video.upload, video.stop_recording, video.clear],
         fn=event_handler_video,
         inputs=[video],
-        outputs=[clear, submit, text],
+        outputs=[
+            clear,
+            submit,
+            text,
+            waveform,
+            faces,
+            emotion_stats,
+            sent_stats,
+            time_row,
+            video_duration,
+            calculate_time,
+        ],
         queue=True,
     )
 

app/event_handlers/submit.py

@@ -55,7 +55,7 @@ def event_handler_submit(
     gr.Textbox,
     gr.Textbox,
 ]:
-    with Timer() as t:
+    with Timer() as timer:
         if video:
             if video.split(".")[-1] == "webm":
                 video = convert_webm_to_mp4(video)
@@ -197,8 +197,15 @@ def event_handler_submit(
         gr.Plot(value=plt_sent, visible=True),
         gr.Row(visible=True),
         gr.Textbox(
-            value=config_data.InformationMessages_VIDEO_DURATION.format(vfe.dur),
+            value=config_data.OtherMessages_SEC.format(vfe.dur),
+            info=config_data.InformationMessages_VIDEO_DURATION,
+            container=True,
+            visible=True,
+        ),
+        gr.Textbox(
+            value=timer.execution_time,
+            info=config_data.InformationMessages_INFERENCE_TIME,
+            container=True,
             visible=True,
         ),
-        gr.Textbox(value=t, visible=True),
     )

app/event_handlers/video.py

@@ -11,7 +11,18 @@ import gradio as gr
 from app.config import config_data
 
 
-def event_handler_video(video: str) -> tuple[gr.Button, gr.Button, gr.Textbox]:
+def event_handler_video(video: str) -> tuple[
+    gr.Button,
+    gr.Button,
+    gr.Textbox,
+    gr.Plot,
+    gr.Plot,
+    gr.Plot,
+    gr.Plot,
+    gr.Row,
+    gr.Textbox,
+    gr.Textbox,
+]:
     is_video_valid = bool(video)
 
     return (
@@ -23,4 +34,11 @@ def event_handler_video(video: str) -> tuple[gr.Button, gr.Button, gr.Textbox]:
             container=False,
             elem_classes="noti-results-" + str(is_video_valid).lower(),
         ),
+        gr.Plot(value=None, visible=False),
+        gr.Plot(value=None, visible=False),
+        gr.Plot(value=None, visible=False),
+        gr.Plot(value=None, visible=False),
+        gr.Row(visible=False),
+        gr.Textbox(value=None, info=None, container=False, visible=False),
+        gr.Textbox(value=None, info=None, container=False, visible=False),
     )

app/tabs.py

@@ -183,12 +183,12 @@ def app_tab():
     )
 
 
-def settings_app_tab():
-    pass
+# def settings_app_tab():
+#     pass
 
 
-def about_app_tab():
-    pass
+# def about_app_tab():
+#     pass
 
 
 def about_authors_tab():

app/utils.py

@@ -32,7 +32,7 @@ class Timer(ContextDecorator):
 
     def __exit__(self, *args):
         self.end = time.time()
-        self.execution_time = f"Inference time: {self.end - self.start:.2f} seconds"
+        self.execution_time = f"{self.end - self.start:.2f} seconds"
 
     def __str__(self):
         return self.execution_time

config.toml

@@ -30,11 +30,13 @@ NOTI_RESULTS = [
     "Video uploaded, you can perform calculations",
 ]
 REC_TEXT = "Recognized text"
-VIDEO_DURATION = "Video duration: {:.2f} seconds"
+VIDEO_DURATION = "Video duration"
+INFERENCE_TIME = "Inference time"
 
 [OtherMessages]
 CLEAR = "Clear"
 SUBMIT = "Calculate"
+SEC = "{:.2f} seconds"
 
 [Labels]
 VIDEO = "Video"
@@ -45,8 +47,8 @@ SENT_STATS = "Statistics of sentiments"
 
 [TabCreators]
 "⭐ App" = "app_tab"
-"⚙️ Settings" = "settings_app_tab"
-"💡 About App" = "about_app_tab"
+# "⚙️ Settings" = "settings_app_tab"
+# "💡 About App" = "about_app_tab"
 "🎭 Authors" = "about_authors_tab"
 "📋 Requirements" = "requirements_app_tab"