App.py

import gradio as gr
from pytube import YouTube
import whisper
from transformers import pipeline
from bark import generate_audio, preload_models
import os
import moviepy.editor as mp

# Load models
asr_model = whisper.load_model("base")
translator = pipeline("translation", model="Helsinki-NLP/opus-mt-en-hi")

# Preload Bark models
preload_models()

def process_video(link):
try:
yt = YouTube(link)
stream = yt.streams.filter(only_audio=True).first()
file_path = stream.download(filename="audio.mp4")

# Step 2: Transcribe English
result = asr_model.transcribe(file_path)
english_text = result["text"]

# Step 3: Translate to Hindi
hindi_text = translator(english_text)[0]['translation_text']

# Step 4: Generate Hindi Audio using Bark
hindi_audio = generate_audio(hindi_text)
with open("hindi.wav", "wb") as f:
f.write(hindi_audio)

# Step 5: Merge audio with video (optional)
original = mp.AudioFileClip("hindi.wav")
video = mp.VideoFileClip(file_path.replace(".mp4", ".mp4")).set_audio(original)
output_path = "dubbed_video.mp4"
video.write_videofile(output_path)

return hindi_text, output_path
except Exception as e:
return str(e), None

gr.Interface(
fn=process_video,
inputs=gr.Textbox(label="Enter YouTube Video URL"),
outputs=[
gr.Textbox(label="Hindi Translation"),
gr.Video(label="Dubbed Hindi Video")
]
).launch()

requirements.txt

@@ -0,0 +1,8 @@
+transformers
+torch
+pytube
+openai-whisper
+moviepy
+gradio
+ffmpeg-python
+git+https://github.com/suno-ai/bark.git