Spaces:
Sleeping
Sleeping
File size: 3,799 Bytes
89b078e 9d2b048 89b078e 9d2b048 89b078e 9d2b048 89b078e 3f75f59 9d2b048 89b078e ade57f6 89b078e 86ce790 89b078e d4a3868 ac669c3 d4a3868 ac669c3 89b078e 9d2b048 0798764 89b078e 9d2b048 0798764 9d2b048 dcad378 89b078e ac669c3 89b078e dcad378 89b078e 19092a8 89b078e 47cbb99 9d2b048 89b078e 0798764 89b078e dcad378 89b078e 9d2b048 47cbb99 9d2b048 |
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 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 |
import streamlit as st
from transformers import pipeline
from pytube import YouTube
from pydub import AudioSegment
from audio_extract import extract_audio
import google.generativeai as google_genai
import os
from dotenv import load_dotenv
load_dotenv()
GOOGLE_API_KEY = os.getenv("GOOGLE_API_KEY")
google_genai.configure(api_key=GOOGLE_API_KEY)
st.set_page_config(
page_title="VidText"
)
def youtube_video_downloader(url):
yt_vid = YouTube(url)
title = yt_vid.title
vid_dld = (
yt_vid.streams.filter(progressive=True, file_extension="mp4")
.order_by("resolution")
.desc()
.first()
)
vid_dld = vid_dld.download()
return vid_dld, title
def audio_extraction(video_file, output_format):
# temp_filename = video_file.name
# video_path = f"{temp_filename}"
audio = extract_audio(
input_path=os.fspath(video_file), output_path=f"{str(video_file)[:-4]}.mp3", output_format=f"{output_format}"
)
return audio
def audio_processing(mp3_audio):
audio = AudioSegment.from_file(mp3_audio, format="mp3")
wav_file = "audio_file.wav"
audio = audio.export(wav_file, format="wav")
return wav_file
@st.cache_resource
def load_model():
asr_model = pipeline(task="automatic-speech-recognition", model="openai/whisper-large-v3")
return asr_model
def transcribe_video(_processed_audio):
transcriber_model = load_model()
text_extract = transcriber_model(_processed_audio)
return text_extract['text']
def generate_ai_summary(transcript):
model = google_genai.GenerativeModel('gemini-pro')
model_response = model.generate_content([f"Give a summary of the text {transcript}"], stream=True)
return model_response.text
# Streamlit UI
youtube_url_tab, file_select_tab, audio_file_tab = st.tabs(
["Youtube url", "Video file", "Audio file"]
)
with youtube_url_tab:
url = st.text_input("Enter the Youtube url")
try:
yt_video, title = youtube_video_downloader(url)
if url:
if st.button("Transcribe", key="yturl"):
with st.spinner("Transcribing..."):
ytvideo_transcript = transcribe_video(yt_video)
st.success(f"Transcription successful")
st.write(ytvideo_transcript)
if st.button("Generate Summary"):
summary = generate_ai_summary(ytvideo_transcript)
st.write(summary)
except Exception as e:
st.error(e)
# Video file transcription
with file_select_tab:
video_file = st.file_uploader("Upload video file", type="mp4")
if video_file:
if st.button("Transcribe", key="vidfile"):
with st.spinner("Transcribing..."):
audio = audio_extraction(os.fspath(video_file), "mp3")
video_transcript = transcribe_video(audio)
st.success(f"Transcription successful")
st.write(video_transcript)
if st.button("Generate Summary", key="ti2"):
summary = generate_ai_summary(video_transcript)
st.write(summary)
# Audio transcription
with audio_file_tab:
audio_file = st.file_uploader("Upload audio file", type="mp3")
if audio_file:
if st.button("Transcribe", key="audiofile"):
with st.spinner("Transcribing..."):
processed_audio = audio_processing(audio_file)
audio_transcript = transcribe_video(processed_audio)
st.success(f"Transcription successful")
st.write(audio_transcript)
if st.button("Generate Summary", key="ti1"):
summary = generate_ai_summary(audio_transcript)
st.write(summary)
|