Update app.py

app.py

@@ -157,89 +157,6 @@ def resample_waveform(waveform, original_sample_rate, target_sample_rate):
 
 
 
-# def split_audio(waveform, sample_rate):
-#     segment_samples = segment_duration * sample_rate
-#     total_samples = waveform.size(0)
-    
-#     segments = []
-#     # If the audio is shorter than the segment duration, just use the entire audio
-#     if total_samples <= segment_samples:
-#         segments.append(waveform)
-#     else:
-#         # Split the audio into segments of the specified duration
-#         for start in range(0, total_samples, segment_samples):
-#             end = min(start + segment_samples, total_samples)
-#             segment = waveform[start:end]
-#             segments.append(segment)
-    
-#     # Ensure we have at least one segment with a minimum length
-#     if len(segments) == 0 or all(len(segment) < 100 for segment in segments):
-#         # Create a padded segment if audio is too short
-#         padded_segment = torch.zeros(segment_samples)
-#         if total_samples > 0:
-#             padded_segment[:total_samples] = waveform
-#         segments = [padded_segment]
-        
-#     return segments
-
-# def split_audio(waveform, sample_rate):
-#     segment_samples = segment_duration * sample_rate
-#     total_samples = waveform.size(0)
-
-#     segments = []
-#     # If the audio is shorter than the segment duration, just use the entire audio
-#     if total_samples <= segment_samples:
-#         segments.append(waveform)
-#     else:
-#         # Split the audio into segments of the specified duration
-#         for start in range(0, total_samples, segment_samples):
-#             end = min(start + segment_samples, total_samples)
-#             segment = waveform[start:end]
-#             segments.append(segment)
-    
-#     # Ensure we have at least one segment
-#     if len(segments) == 0:
-#         segments.append(waveform)
-
-#     return segments
-
-# def split_audio(waveform, sample_rate):
-#     segment_samples = segment_duration * sample_rate
-#     total_samples = waveform.size(0)
-
-#     # Pad if shorter than one segment
-#     if total_samples < segment_samples:
-#         pad_size = segment_samples - total_samples
-#         waveform = torch.nn.functional.pad(waveform, (0, pad_size))
-
-#     segments = []
-#     for start in range(0, waveform.size(0), segment_samples):
-#         end = start + segment_samples
-#         if end <= waveform.size(0):
-#             segment = waveform[start:end]
-#             segments.append(segment)
-
-#     return segments
-
-# def split_audio(waveform, sample_rate, segment_duration=10):
-#     segment_samples = segment_duration * sample_rate
-#     total_samples = waveform.size(0)
-
-#     segments = []
-#     for start in range(0, total_samples, segment_samples):
-#         end = start + segment_samples
-#         if end <= total_samples:
-#             segment = waveform[start:end]
-#             segments.append(segment)
-
-#     # If no full segments were created, pad the short waveform
-#     if len(segments) == 0:
-#         pad_length = segment_samples - total_samples
-#         padded_waveform = torch.nn.functional.pad(waveform, (0, pad_length))
-#         segments.append(padded_waveform)
-
-#     return segments
-
 def split_audio(waveform, sample_rate):
     segment_samples = segment_duration * sample_rate
     total_samples = waveform.size(0)
@@ -778,13 +695,12 @@ css = """
 
 def analyze_emotions(audio_path, threshold):
     if audio_path is None:
-        return "❌ Please upload a valid audio file.", None, None
+        return {"error": "Please upload a valid audio file"}
     try:
         model_output = music2emo.predict(audio_path, threshold)
         return model_output
     except Exception as e:
-        return f"⚠️ Error during prediction: {e}", None, None
-
+        return {"error": str(e)}
 
 with gr.Blocks(css=css) as demo:
     gr.HTML(f"<h1 style='text-align: center;'>{title}</h1>")
@@ -816,35 +732,13 @@ with gr.Blocks(css=css) as demo:
         
         # Right Panel (Output)
         with gr.Column(scale=1):
-            output_text = gr.Textbox(
-                label="Analysis Results",
-                lines=4,
-                interactive=False  # Prevent user input
-            )
-
-            # Ensure both plots have padding on top
-            with gr.Row(equal_height=True):
-                mood_chart = gr.Plot(label="Mood Probabilities", scale=2, elem_classes=["gr-box"])
-                va_chart = gr.Plot(label="Valence-Arousal Space", scale=1, elem_classes=["gr-box"])
+            output_json = gr.JSON(label="Model Output")
 
-    # predict_btn.click(
-    #     fn=lambda audio, thresh: format_prediction(music2emo.predict(audio, thresh)),
-    #     inputs=[input_audio, threshold],
-    #     outputs=[output_text, va_chart, mood_chart]
-    # )
     predict_btn.click(
-        fn=lambda audio, thresh: format_prediction(music2emo.predict(audio, thresh)),
+        fn=analyze_emotions,
         inputs=[input_audio, threshold],
-        outputs=[output_text, va_chart, mood_chart]
+        outputs=output_json
     )
 
-    demo.load(fn=analyze_emotions, inputs=[input_audio, threshold], outputs=[model_output])
-
 # Launch the App
-demo.queue().launch(share=True, show_error=True)
-
-
-
-
-
-    
+demo.queue().launch(share=True, show_error=True)