Spaces:
Build error
Build error
| import streamlit as st | |
| import youtube_dl | |
| from moviepy.editor import VideoFileClip | |
| import speech_recognition as sr | |
| from transformers import pipeline | |
| # Initialize the Hugging Face text generation pipeline (using GPT-Neo or GPT-2) | |
| generator = pipeline("text-generation", model="EleutherAI/gpt-neo-2.7B") | |
| def download_video(url): | |
| ydl_opts = { | |
| 'format': 'bestvideo+bestaudio/best', | |
| 'outtmpl': 'downloaded_video.mp4' | |
| } | |
| with youtube_dl.YoutubeDL(ydl_opts) as ydl: | |
| ydl.download([url]) | |
| def trim_video(video_path, start_time, end_time): | |
| video = VideoFileClip(video_path) | |
| trimmed_video = video.subclip(start_time, end_time) | |
| trimmed_video.write_videofile("trimmed_video.mp4") | |
| def transcribe_audio(video_path): | |
| recognizer = sr.Recognizer() | |
| with sr.AudioFile(video_path) as source: | |
| audio = recognizer.record(source) | |
| try: | |
| text = recognizer.recognize_google(audio) # This uses Google's free Web Speech API | |
| return text | |
| except sr.UnknownValueError: | |
| return "Sorry, I could not understand the audio." | |
| except sr.RequestError: | |
| return "Error with the API request." | |
| def generate_subtitles(text): | |
| # Use Hugging Face's GPT-Neo model for text generation | |
| generated_text = generator(text, max_length=150, num_return_sequences=1) | |
| return generated_text[0]['generated_text'] | |
| # Streamlit Interface | |
| st.title("YouTube Video Trimmer and Subtitle Generator") | |
| url = st.text_input("Enter YouTube URL") | |
| start_time = st.selectbox("Select Start Time", [0, 15, 30, 60]) | |
| end_time = st.selectbox("Select End Time", [15, 30, 60, 90]) | |
| font_style = st.selectbox("Select Font Style", ["Arial", "Helvetica", "Times New Roman", "Courier"]) | |
| if st.button("Generate"): | |
| # Download and trim the video | |
| download_video(url) | |
| trim_video("downloaded_video.mp4", start_time, end_time) | |
| # Transcribe audio to text | |
| transcription = transcribe_audio("trimmed_video.mp4") | |
| # Generate subtitles using Hugging Face's GPT model | |
| subtitles = generate_subtitles(transcription) | |
| # Show trimmed video and generated subtitles | |
| st.video("trimmed_video.mp4") | |
| st.subheader("Subtitles:") | |
| st.write(subtitles) | |