app4

README.md

@@ -5,7 +5,7 @@ colorFrom: green
 colorTo: indigo
 sdk: gradio
 sdk_version: 3.0.19
-app_file: app3.py
+app_file: app4.py
 pinned: false
 duplicated_from: harish3110/emotion_detection
 ---

app4.py

@@ -0,0 +1,63 @@
+from speechbrain.pretrained.interfaces import foreign_class
+import gradio as gr
+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", 
+    classname="CustomEncoderWav2vec2Classifier"
+)
+
+# Building prediction function for Gradio
+emotion_dict = {
+    'sad': 'Sad', 
+    'hap': 'Happy',
+    'ang': 'Anger',
+    'fea': 'Fear',
+    'sur': 'Surprised',
+    'neu': 'Neutral'
+}
+
+def predict_emotion(selected_audio):
+    if selected_audio is None:  # Check if an audio file is selected
+        return "Please select an audio file.", None
+    file_path = os.path.join("rec", selected_audio)
+    out_prob, score, index, text_lab = learner.classify_file(file_path)
+    emotion = emotion_dict[text_lab[0]]
+    return emotion, file_path  # Return both emotion and file path
+
+# Get the list of audio files for the dropdown
+audio_files_list = get_audio_files_list()
+
+# Loading Gradio interface
+dropdown = gr.Dropdown(label="Select Audio", choices=audio_files_list)
+button = gr.Button("Detect emotion")
+outputs = [gr.outputs.Textbox(label="Predicted Emotion"), gr.outputs.Audio(label="Play Audio")]
+
+def button_click(selected_audio):
+    return predict_emotion(selected_audio)  # Call predict_emotion when button is clicked
+
+title = "ML Speech Emotion Detection"
+description = "Speechbrain powered wav2vec 2.0 pretrained model on IEMOCAP dataset using Gradio."
+
+# Create the Gradio interface
+interface = gr.Interface(
+    fn=button_click,  # Use the button_click function for the interface
+    inputs=[dropdown, button],
+    outputs=outputs,
+    title=title,
+    description=description
+)
+
+interface.launch()