Spaces:
Runtime error
Runtime error
| 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/' directory1 | |
| 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): | |
| file_path = os.path.join("rec", selected_audio) | |
| out_prob, score, index, text_lab = learner.classify_file(file_path) | |
| return emotion_dict[text_lab[0]] | |
| # Get the list of audio files for the dropdown | |
| audio_files_list = get_audio_files_list() | |
| # Function to return the selected audio file path | |
| def get_audio_file_path(selected_audio): | |
| file_path = os.path.join("rec", selected_audio) | |
| return file_path | |
| # Gradio components | |
| dropdown = gr.Dropdown(label="Select Audio", choices=audio_files_list) | |
| audio_player = gr.Audio(source="file", label="Listen to the selected audio") | |
| # Update the audio player when a new selection is made from the dropdown | |
| def update_audio(selected_audio): | |
| return get_audio_file_path(selected_audio) | |
| # Connect the dropdown to the audio player using the update_audio function | |
| dropdown.change(fn=update_audio, inputs=dropdown, outputs=audio_player) | |
| # Update the Gradio interface to use both the dropdown and the audio player as inputs | |
| interface = gr.Interface( | |
| fn=predict_emotion, | |
| inputs=[dropdown, audio_player], | |
| outputs="text", | |
| title="ML Speech Emotion Detection", | |
| description="Speechbrain powered wav2vec 2.0 pretrained model on IEMOCAP dataset using Gradio." | |
| ) | |
| interface.launch() |