Update app.py

app.py

@@ -18,10 +18,10 @@ def generate_mel_spectrogram(audio_path, sr=22050, n_mels=128, fmin=0, fmax=7000
     
     return S_dB, y, sr
 
-def detect_zero_db(spectrogram):
+def detect_zero_db(spectrogram,threshold,tol):
     # Create a binary mask where the spectrogram values are close to 0 dB
-    threshold = -40  # +0 dB threshold
-    mask = np.isclose(spectrogram, threshold, atol=10)  # Use a tolerance to include values close to 0 dB
+    # +0 dB threshold
+    mask = np.isclose(spectrogram, threshold, atol=tol)  # Use a tolerance to include values close to 0 dB
     
     return mask
 
@@ -53,9 +53,9 @@ def plot_frequency(times, frequencies, label, color, file_path):
     plt.savefig(file_path, format='png', bbox_inches='tight', pad_inches=0)
     plt.close()
 
-def process_audio(audio_file):
+def process_audio( threshold, audio_file,tol):
     mel_spectrogram, y, sr = generate_mel_spectrogram(audio_file)
-    edges = detect_zero_db(mel_spectrogram)
+    edges = detect_zero_db(mel_spectrogram,threshold,tol)
 
     # Create temporary files to save the generated images
     with tempfile.NamedTemporaryFile(suffix=".png", delete=False) as mel_file, \
@@ -110,13 +110,17 @@ def process_audio(audio_file):
 
 with gr.Blocks() as demo:
     with gr.Group():
+        threshold_slider =gr.Slider(-100,0,value=-20,info="Choose between -100 and 0")
+        tol_slider =gr.Slider(0,25,value=10,info="Choose between 0 and 25")
         audio_input = gr.Audio(label="Upload an audio file in WAV format", type="filepath")
+        submit_button = gr.Button("Submit")
         img_slider = ImageSlider(label="Before and After Edge Detection", type="filepath", slider_color="pink")
         f0_plot = gr.Image(label="F0 Frequency Plot", type="filepath")
         f1_plot = gr.Image(label="F1 Frequency Plot", type="filepath")
         f2_plot = gr.Image(label="F2 Frequency Plot", type="filepath")
         
-        audio_input.upload(process_audio, inputs=audio_input, outputs=[img_slider, f0_plot, f1_plot, f2_plot])
+        
+        submit_button.click(process_audio, inputs=[ threshold_slider, audio_input,tol_slider], outputs=[img_slider, f0_plot, f1_plot, f2_plot])
 
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch()