File size: 2,804 Bytes
2a284c3 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 |
# This Gradio app demonstrates a local voice cloning demo using the provided video and voice samples.
import gradio as gr
import numpy as np
import tempfile
import os
# Function to clone voice and generate dubbed video
def clone_voice(video_path, voice_option):
# Create a temporary directory to store intermediate files
temp_dir = tempfile.mkdtemp()
audio_path = os.path.join(temp_dir, "original_audio.wav")
output_path = os.path.join(temp_dir, "output_video.mp4")
# 1. Extract audio from the uploaded video
# Note: Since moviepy is not available, we will use a placeholder function to simulate audio extraction
# In a real scenario, you would use a library like moviepy to extract audio from the video
with open(video_path, 'rb') as video_file:
video_data = video_file.read()
with open(audio_path, 'wb') as audio_file:
audio_file.write(video_data) # Placeholder for audio extraction
# 2. Transcribe the extracted audio to text using a placeholder function
# Note: Since the Whisper model is not available, we will use a placeholder function to simulate transcription
# In a real scenario, you would use a transcription model like Whisper
text = "This is a placeholder text for the transcribed audio."
# 3. Generate the cloned voice using the TTS model
# Note: Since TTS is not available, we will use a placeholder function to simulate TTS
# In a real scenario, you would use a TTS library to generate the cloned voice
with open(os.path.join(temp_dir, "new_audio.wav"), 'wb') as new_audio_file:
new_audio_file.write(text.encode()) # Placeholder for TTS
# 4. Replace the original audio in the video with the new cloned audio
# Note: Since moviepy is not available, we will use a placeholder function to simulate video creation
# In a real scenario, you would use a library like moviepy to create the final video
with open(output_path, 'wb') as final_video_file:
final_video_file.write(video_data) # Placeholder for video creation
return output_path
# Create the Gradio interface
with gr.Blocks() as demo:
gr.Markdown("## Local Voice Cloning Demo")
with gr.Row():
video_input = gr.Video(label="Upload Video", sources=["upload"])
voice_selector = gr.Dropdown(
choices=["voice1", "voice2", "voice3"],
label="Select Voice Profile",
info="Add 3-5 second WAV samples in 'voices/' folder"
)
submit_btn = gr.Button("Generate Dubbed Video")
video_output = gr.Video(label="Result")
submit_btn.click(
fn=clone_voice,
inputs=[video_input, voice_selector],
outputs=video_output
)
# Launch the Gradio app
if __name__ == "__main__":
demo.launch(show_error=True) |