🎤 ModalX-AI

ModalX-AI is a lightweight multimodal analysis tool that converts videos into clean, readable transcripts using AI. It supports local videos, direct video links, and YouTube videos, all through a simple Streamlit interface.

🚀 Features (Phase 1)

📤 Upload local video files

🌐 Download videos via direct links

▶️ Experimental YouTube video support

🔊 Automatic audio extraction

🧠 AI-powered speech-to-text transcription (Whisper)

📄 Transcript saved as .txt

💻 Clean Streamlit UI

🧠 Tech Stack

Frontend: Streamlit

Audio/Video: MoviePy + FFmpeg

Transcription: OpenAI Whisper (local, offline)

Downloads: yt-dlp

Language: Python

📁 Project Structure modalX-ai-analysis/ │ ├── app/ │ └── app.py │ ├── core/ │ ├── file_handler.py │ ├── video_downloader.py │ ├── youtube_downloader.py │ ├── video_to_audio.py │ └── transcriber.py │ ├── data/ │ ├── input_videos/ │ ├── extracted_audio/ │ └── transcripts/ │ ├── requirements.txt └── README.md

⚙️ Setup & Run 1️⃣ Install dependencies pip install -r requirements.txt

2️⃣ Install FFmpeg (Required)

Download from: https://www.gyan.dev/ffmpeg/builds/

Add ffmpeg/bin to system PATH

Verify:

ffmpeg -version

3️⃣ Run the app streamlit run app/app.py

📌 Usage Flow

Upload a video / paste link / paste YouTube URL

Click Extract Audio

Click Generate Transcript

Transcript appears on screen and is saved automatically

⚠️ Notes

YouTube support is experimental

CPU-only systems may run slower (normal)

Internet is required only for downloads, not transcription

🧩 Roadmap

Phase 2: Summarization & key insights

Phase 3: Slide & presentation analysis

Phase 4: Export reports (PDF / DOCX)

👨‍💻 Author

Mahadi Rahman Jihad CSE Undergrad | AI & Data Science enthusiast