tuan3w/visual_search
A visual search engine that uses Faster-RCNN CNN feature extraction to enable similarity-based image retrieval.
Not currently ranked — collecting fresh signals.
star history
This project implements a visual search system using deep learning feature extraction. A Faster-RCNN model with VGG-16 backbone extracts feature vectors from images, which are then stored in Elasticsearch with a custom distance-computing plugin. The entire stack is dockerized, with the Python-based feature extraction service communicating with an Elasticsearch container over a Docker network to enable approximate nearest-neighbor image search.
Frequently asked
- What is tuan3w/visual_search?
- A visual search engine that uses Faster-RCNN CNN feature extraction to enable similarity-based image retrieval.
- Is visual_search open source?
- Yes — tuan3w/visual_search is open source, released under the MIT license.
- What language is visual_search written in?
- tuan3w/visual_search is primarily written in JavaScript.
- How popular is visual_search?
- tuan3w/visual_search has 443 stars on GitHub.
- Where can I find visual_search?
- tuan3w/visual_search is on GitHub at https://github.com/tuan3w/visual_search.