|
--- |
|
title: Mutimodal Video Chat RAG |
|
emoji: 🦀 |
|
colorFrom: purple |
|
colorTo: purple |
|
sdk: gradio |
|
sdk_version: 5.3.0 |
|
app_file: app.py |
|
pinned: false |
|
license: mit |
|
short_description: This is a RAG system for chatting with videos |
|
--- |
|
|
|
# Video Chat RAG |
|
|
|
This is a Retrieval-Augmented Generation (RAG) system for chatting with videos. Upload a video and ask questions about its content! |
|
|
|
## Features |
|
|
|
- Video frame analysis using CLIP |
|
- Natural language querying of video content |
|
- Frame retrieval with relevance scores |
|
- User-friendly interface |
|
|
|
## Usage |
|
|
|
1. Upload a video using the file upload button |
|
2. Click "Process Video" and wait for processing to complete |
|
3. Type your question in the text box |
|
4. Click "Search" to find relevant frames |
|
|
|
## Technical Details |
|
|
|
The system uses: |
|
- CLIP for visual-language understanding |
|
- FAISS for efficient similarity search |
|
- Gradio for the user interface |
|
|
|
## Limitations |
|
|
|
- Maximum video file size: 500MB |
|
- Supported formats: MP4 |
|
- Processing time depends on video length |
|
|
