https://github.com/bwconrad/video-content-search

Search the content of a video with a text or image query
https://github.com/bwconrad/video-content-search

computer-vision deep-learning video-search

Last synced: about 1 month ago
JSON representation

Search the content of a video with a text or image query

• Host: GitHub
• URL: https://github.com/bwconrad/video-content-search
• Owner: bwconrad
• License: mit
• Created: 2023-04-17T19:40:09.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2023-04-20T19:25:31.000Z (over 1 year ago)
• Last Synced: 2024-05-14T00:18:10.784Z (8 months ago)
• Topics: computer-vision, deep-learning, video-search
• Language: Jupyter Notebook
• Homepage:
• Size: 1.89 MB
• Stars: 4
• Watchers: 1
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Video Content Search

[![Hugging Face Spaces](https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Spaces-blue)](https://huggingface.co/spaces/bwconrad/video-content-search)

An application for searching videos to find frames that match a text or image query. This application utilizes ConvNext CLIP models from [OpenCLIP](https://github.com/mlfoundations/open_clip) to compare video frames with the feature representation of a user text or image query.









## Requirements

- Python 3.8+

- `pip install -r requirements`

## Usage

#### Web App

The application is available in the Hugging Face Space [here](https://huggingface.co/spaces/bwconrad/video-content-search).

Or you can run it locally by doing the following:

1. Run `python video-search-app.py` and open the given URL in a browser. 

2. Select either the "Text Query Search" or "Image Query Search" tab.

3. Upload your video and write a text query or upload a query image.

4. Adjust the parameters and click submit.

- __Note__: The web app can have performance issues with long videos (more than a few minutes). Instead use the notebook for longer videos.

#### Notebook

[`video-search-notebook.ipynb`](video-search-notebook.ipynb) provides an alternate UI with a few extra features.