Update app.py

app.py

@@ -1,20 +1,25 @@
-from speechbrain.pretrained.interfaces import foreign_class
 import gradio as gr
+from speechbrain.pretrained.interfaces import foreign_class
 import os
 
-import warnings
-warnings.filterwarnings("ignore")
+# Function to get the list of audio files in the 'rec/' directory
+def get_audio_files_list(directory="rec"):
+    try:
+        return [f for f in os.listdir(directory) if os.path.isfile(os.path.join(directory, f))]
+    except FileNotFoundError:
+        print("The 'rec' directory does not exist. Please make sure it is the correct path.")
+        return []
 
 # Loading the speechbrain emotion detection model
 learner = foreign_class(
     source="speechbrain/emotion-recognition-wav2vec2-IEMOCAP",
-    pymodule_file="custom_interface.py", 
+    pymodule_file="custom_interface.py",
     classname="CustomEncoderWav2vec2Classifier"
 )
 
-# Building prediction function for gradio
+# Building prediction function for Gradio
 emotion_dict = {
-    'sad': 'Sad', 
+    'sad': 'Sad',
     'hap': 'Happy',
     'ang': 'Anger',
     'fea': 'Fear',
@@ -22,24 +27,29 @@ emotion_dict = {
     'neu': 'Neutral'
 }
 
-
-
-# Assuming emotion_dict and learner are defined elsewhere in your code
-# and learner.classify_file is a method that classifies the audio file
-
-def predict_emotion(audio, rec_file):
-    rec_path = os.path.join("rec", rec_file.name)
-    # Assuming you want to use the audio file from the 'rec' directory for prediction
-    out_prob, score, index, text_lab = learner.classify_file(rec_path)
-    return emotion_dict[text_lab[0]]
-
-# Loading gradio interface
-inputs = [
-    gr.inputs.Audio(label="Input Audio", type="file"),
-    gr.inputs.File(label="Choose file from rec directory", type="file", default="rec/")
-]
-outputs = "text"
-title = "ML Speech Emotion Detection"
-description = "Speechbrain powered wav2vec 2.0 pretrained model on IEMOCAP dataset using Gradio."
-
-gr.Interface(fn=predict_emotion, inputs=inputs, outputs=outputs, title=title, description=description).launch()
+def selected_audio(audio_file):
+    if audio_file is None:
+        return None, "Please select an audio file."
+    file_path = os.path.join("rec", audio_file)
+    audio_data = gr.Audio(file=file_path)
+    out_prob, score, index, text_lab = learner.classify_file(file_path)
+    emotion = emotion_dict[text_lab[0]]
+    return audio_data, emotion
+
+# Get the list of audio files for the dropdown
+audio_files_list = get_audio_files_list()
+
+# Define Gradio blocks
+with gr.Blocks() as blocks:
+    gr.Markdown("<h1 style='text-align: center; margin-bottom: 1rem'>" +
+                "Audio Emotion Detection" +
+                "</h1>")
+    with gr.Column():
+        input_audio_dropdown = gr.Dropdown(label="Select Audio", choices=audio_files_list)
+        audio_ui = gr.Audio()
+        output_text = gr.Textbox(label="Detected Emotion!")
+        detect_btn = gr.Button("Detect Emotion")
+        detect_btn.click(selected_audio, inputs=input_audio_dropdown, outputs=[audio_ui, output_text])
+
+# Launch the Gradio blocks interface
+blocks.launch()