Update app.py

app.py

@@ -4,7 +4,9 @@ from processing import process_input
 from transcription_diarization import diarize_audio
 from visualization import create_charts
 import time
+import re
 import cv2
+import os
 from config import openai_api_key
 
 # Load the model
@@ -13,7 +15,7 @@ llm = load_model(openai_api_key)
 def analyze_video(video_path, progress=gr.Progress()):
     start_time = time.time()
     if not video_path:
-        return [None] * 32  # Return None for all outputs
+        return [None] * 29  # Return None for all outputs
 
     progress(0, desc="Starting analysis...")
     progress(0.2, desc="Starting transcription and diarization")
@@ -32,41 +34,91 @@ def analyze_video(video_path, progress=gr.Progress()):
     end_time = time.time()
     execution_time = end_time - start_time
 
-    output_components = []
+    output_components = []  # transcript
 
     output_components.append(f"Completed in {int(execution_time)} seconds.")
     output_components.append(gr.Textbox(value=transcription, label="Transcript", lines=10, visible=True))
 
-    for i in range(3):  # Always process 3 speakers, even if some are empty
-        speaker_id = f"Speaker {i+1}"
-        speaker_charts = charts.get(speaker_id, {})
-        speaker_explanations = explanations.get(speaker_id, {})
-        speaker_general_impression = general_impressions.get(speaker_id, "")
+        with gr.Tab(label=f'Examples', visible=False):
+        gr.Markdown("### Example Videos")
+        with gr.Row():
+            with gr.Column(scale=1):
+                example_video_1_path = "examples/Scenes.From.A.Marriage.US.mp4"
+                preview_1 = get_middle_frame(example_video_1_path)
+                gr.Image(preview_1, label="Scenes From A Marriage")
+                example_video_1 = gr.Video(example_video_1_path, label="Example 1", visible=False)
+                use_example_button_1 = gr.Button("Load Example 1")
+            
+            with gr.Column(scale=1):
+                example_video_2_path = "examples/Billie Eilish.mp4"
+                preview_2 = get_middle_frame(example_video_2_path)
+                gr.Image(preview_2, label="Billie Eilish")
+                example_video_2 = gr.Video(example_video_2_path, label="Example 2", visible=False)
+                use_example_button_2 = gr.Button("Load Example 2")
+            
+            with gr.Column(scale=1):
+                example_video_3_path = "examples/Elliot Rodger.mp4"
+                preview_3 = get_middle_frame(example_video_3_path)
+                gr.Image(preview_3, label="Elliot Rodger")
+                example_video_3 = gr.Video(example_video_3_path, label="Example 3", visible=False)
+                use_example_button_3 = gr.Button("Load Example 3")
+
+    with gr.Tab(label=f'Description', visible=False):
+        gr.Markdown(description_txt)
+        gr.HTML("<div style='height: 20px;'></div>")
+        gr.Image(value="appendix/AI Personality Detection flow - 1.png", label='Flowchart 1', width=1000)
+        gr.Image(value="appendix/AI Personality Detection flow - 2.png", label='Flowchart 2', width=1000)
+
+    for i, (speaker_id, speaker_charts) in enumerate(charts.items(), start=1):
+        print(speaker_id)
+        speaker_explanations = explanations[speaker_id]
+        speaker_general_impression = general_impressions[speaker_id]
         
-        output_components.extend([
-            gr.Markdown(f"### {speaker_id}", visible=True),
-            gr.Textbox(value=speaker_general_impression, label="General Impression", visible=True),
-            gr.Plot(value=speaker_charts.get("attachment", None), visible=True),
-            gr.Plot(value=speaker_charts.get("dimensions", None), visible=True),
-            gr.Textbox(value=speaker_explanations.get("attachment", ""), label="Attachment Styles Explanation", visible=True),
-            gr.Plot(value=speaker_charts.get("bigfive", None), visible=True),
-            gr.Textbox(value=speaker_explanations.get("bigfive", ""), label="Big Five Traits Explanation", visible=True),
-            gr.Plot(value=speaker_charts.get("personality", None), visible=bool(speaker_charts)),
-            gr.Textbox(value=speaker_explanations.get("personality", ""), label="Personality Disorders Explanation", visible=True)
-        ])
-
-    # Ensure we always return 32 components
-    while len(output_components) < 32:
-        output_components.append(None)
+        with gr.Tab():
+            with gr.TabItem(label=f'General Impression'):
+                speaker_section1 = [
+                    gr.Markdown(f"### {speaker_id}", visible=True),
+                    gr.Textbox(value=speaker_general_impression, label="General Impression", visible=True, lines=10)
+                ]
+
+            with gr.TabItem(label=f'Attachment Styles'):
+                speaker_section2 = [
+                    gr.Plot(value=speaker_charts.get("attachment", None), visible=True),
+                    gr.Plot(value=speaker_charts.get("dimensions", None), visible=True),
+                    gr.Textbox(value=speaker_explanations.get("attachment", ""), label="Attachment Styles Explanation",
+                               visible=True, lines=2)
+                ]
+
+            with gr.TabItem(label=f'Big Five Traits'):
+                speaker_section3 = [
+                    gr.Plot(value=speaker_charts.get("bigfive", None), visible=True),
+                    gr.Textbox(value=speaker_explanations.get("bigfive", ""), label="Big Five Traits Explanation",
+                               visible=True, lines=2)
+                ]
+
+            with gr.TabItem(label=f'Personalities'):
+                speaker_section4 = [
+                    gr.Plot(value=speaker_charts.get("personality", None), visible=True),
+                    gr.Textbox(value=speaker_explanations.get("personality", ""),
+                               label="Personality Disorders Explanation", visible=True, lines=2)
+                ]
+
+        output_components.extend(speaker_section1)
+        output_components.extend(speaker_section2)
+        output_components.extend(speaker_section3)
+        output_components.extend(speaker_section4)
+
+    # Pad with None for any missing speakers
+    while len(output_components) < 28:
+        output_components.extend([gr.update(visible=False)] * 9)
 
     return output_components
 
+
 def use_example_1():
     return "examples/Scenes.From.A.Marriage.US.mp4"
-
 def use_example_2():
     return "examples/Billie Eilish.mp4"
-
 def use_example_3():
     return "examples/Elliot Rodger.mp4"
 
@@ -83,9 +135,6 @@ def get_middle_frame(video_path):
         return preview_path
     return None
 
-def clear_outputs():
-    return [None] * 32
-
 with gr.Blocks() as iface:
     gr.Markdown("# Multiple-Speakers-Personality-Analyzer")
     gr.Markdown("This project provides an advanced AI system designed for diagnosing and profiling personality attributes from video content based on a single speaker or multiple speakers in a conversation.")
@@ -97,48 +146,65 @@ with gr.Blocks() as iface:
             
     # Create output components
     output_components = []
+    # Add transcript output near the top
     execution_box = gr.Textbox(label="Execution Info", value="N/A", lines=1)
     output_components.append(execution_box)
 
     transcript = gr.Textbox(label="Transcript", lines=10, visible=False)
     output_components.append(transcript)
 
-    for n in range(3):  # Always create 3 speaker tabs
-        with gr.Tab(label=f'Speaker {n + 1}', visible=True) as tab:
-            gr.Markdown(visible=False)
-            gr.Textbox(label="General Impression", visible=False)
-            gr.Plot(visible=False)
-            gr.Plot(visible=False)
-            gr.Textbox(label="Attachment Styles Explanation", visible=False)
-            gr.Plot(visible=False)
-            gr.Textbox(label="Big Five Traits Explanation", visible=False)
-            gr.Plot(visible=False)
-            gr.Textbox(label="Personality Disorders Explanation", visible=False)
-            output_components.extend([tab] + [component for component in tab.children])
-
-
-    gr.Markdown("### Example Videos")
-    with gr.Row():
-        with gr.Column(scale=1):
-            example_video_1_path = "examples/Scenes.From.A.Marriage.US.mp4"
-            preview_1 = get_middle_frame(example_video_1_path)
-            gr.Image(preview_1, label="Scenes From A Marriage")
-            example_video_1 = gr.Video(example_video_1_path, label="Example 1", visible=False)
-            use_example_button_1 = gr.Button("Load Example 1")
-        
-        with gr.Column(scale=1):
-            example_video_2_path = "examples/Billie Eilish.mp4"
-            preview_2 = get_middle_frame(example_video_2_path)
-            gr.Image(preview_2, label="Billie Eilish")
-            example_video_2 = gr.Video(example_video_2_path, label="Example 2", visible=False)
-            use_example_button_2 = gr.Button("Load Example 2")
-        
-        with gr.Column(scale=1):
-            example_video_3_path = "examples/Elliot Rodger.mp4"
-            preview_3 = get_middle_frame(example_video_3_path)
-            gr.Image(preview_3, label="Elliot Rodger")
-            example_video_3 = gr.Video(example_video_3_path, label="Example 3", visible=False)
-            use_example_button_3 = gr.Button("Load Example 3")
+    for n in range(3):  # Assuming maximum of 3 speakers
+        with gr.Tab(label=f'Speaker {n + 1}', visible=True):
+            with gr.TabItem(label=f'General Impression'):
+                column_components1 = [
+                    gr.Markdown(visible=False),
+                    gr.Textbox(label="General Impression", visible=False)]
+
+            with gr.TabItem(label=f'Attachment Styles'):
+                column_components2 = [
+                    gr.Plot(visible=False),
+                    gr.Plot(visible=False),
+                    gr.Textbox(label="Attachment Styles Explanation", visible=False)]
+
+            with gr.TabItem(label=f'Big Five Traits'):
+                column_components3 = [
+                    gr.Plot(visible=False),
+                    gr.Textbox(label="Big Five Traits Explanation", visible=False)]
+
+            with gr.TabItem(label=f'Personalities'):
+                column_components4 = [
+                    gr.Plot(visible=False),
+                    gr.Textbox(label="Personality Disorders Explanation", visible=False)]
+
+        output_components.extend(column_components1)
+        output_components.extend(column_components2)
+        output_components.extend(column_components3)
+        output_components.extend(column_components4)
+
+    
+    with gr.Tab(label=f'Examples', visible=True):
+        gr.Markdown("### Example Videos")
+        with gr.Row():
+            with gr.Column(scale=1):
+                example_video_1_path = "examples/Scenes.From.A.Marriage.US.mp4"
+                preview_1 = get_middle_frame(example_video_1_path)
+                gr.Image(preview_1, label="Scenes From A Marriage")
+                example_video_1 = gr.Video(example_video_1_path, label="Example 1", visible=False)
+                use_example_button_1 = gr.Button("Load Example 1")
+            
+            with gr.Column(scale=1):
+                example_video_2_path = "examples/Billie Eilish.mp4"
+                preview_2 = get_middle_frame(example_video_2_path)
+                gr.Image(preview_2, label="Billie Eilish")
+                example_video_2 = gr.Video(example_video_2_path, label="Example 2", visible=False)
+                use_example_button_2 = gr.Button("Load Example 2")
+            
+            with gr.Column(scale=1):
+                example_video_3_path = "examples/Elliot Rodger.mp4"
+                preview_3 = get_middle_frame(example_video_3_path)
+                gr.Image(preview_3, label="Elliot Rodger")
+                example_video_3 = gr.Video(example_video_3_path, label="Example 3", visible=False)
+                use_example_button_3 = gr.Button("Load Example 3")
             
     with open('description.txt', 'r') as file:
         description_txt = file.read()
@@ -147,13 +213,8 @@ with gr.Blocks() as iface:
         gr.HTML("<div style='height: 20px;'></div>")
         gr.Image(value="appendix/AI Personality Detection flow - 1.png", label='Flowchart 1', width=1000)
         gr.Image(value="appendix/AI Personality Detection flow - 2.png", label='Flowchart 2', width=1000)
-
             
     analyze_button.click(
-        fn=clear_outputs,
-        inputs=[],
-        outputs=output_components
-    ).then(
         fn=analyze_video,
         inputs=[video_input],
         outputs=output_components,
@@ -164,46 +225,29 @@ with gr.Blocks() as iface:
         fn=use_example_1,
         inputs=[],
         outputs=[video_input],
-    ).then(
-        fn=clear_outputs,
-        inputs=[],
-        outputs=output_components
-    ).then(
-        fn=analyze_video,
-        inputs=[video_input],
-        outputs=output_components,
-        show_progress=True
-    )
-
+    ).then(fn=analyze_video,
+           inputs=[video_input],
+           outputs=output_components,
+           show_progress=True
+           )
     use_example_button_2.click(
         fn=use_example_2,
         inputs=[],
         outputs=[video_input],
-    ).then(
-        fn=clear_outputs,
-        inputs=[],
-        outputs=output_components
-    ).then(
-        fn=analyze_video,
-        inputs=[video_input],
-        outputs=output_components,
-        show_progress=True
-    )
-
+    ).then(fn=analyze_video,
+           inputs=[video_input],
+           outputs=output_components,
+           show_progress=True
+           )
     use_example_button_3.click(
         fn=use_example_3,
         inputs=[],
         outputs=[video_input],
-    ).then(
-        fn=clear_outputs,
-        inputs=[],
-        outputs=output_components
-    ).then(
-        fn=analyze_video,
-        inputs=[video_input],
-        outputs=output_components,
-        show_progress=True
-    )
+    ).then(fn=analyze_video,
+           inputs=[video_input],
+           outputs=output_components,
+           show_progress=True
+           )
 
 if __name__ == "__main__":
     iface.launch()