sfzhang15/RefineDet
Neural network implementation for single-shot object detection published at CVPR 2018.
Not currently ranked — collecting fresh signals.
star history
RefineDet is an object detection model that achieves better accuracy than two-stage methods while maintaining comparable efficiency of one-stage methods. The repository provides code for training and evaluating the detector, implemented in C++, with support for multi-scale testing. It uses VGG16 as the backbone and achieves 81.8% mAP on VOC2007 at 24 FPS on a Titan X GPU.
Frequently asked
- What is sfzhang15/RefineDet?
- Neural network implementation for single-shot object detection published at CVPR 2018.
- Is RefineDet open source?
- Yes — sfzhang15/RefineDet is an open-source project tracked on heatdrop.
- What language is RefineDet written in?
- sfzhang15/RefineDet is primarily written in C++.
- How popular is RefineDet?
- sfzhang15/RefineDet has 1.4k stars on GitHub.
- Where can I find RefineDet?
- sfzhang15/RefineDet is on GitHub at https://github.com/sfzhang15/RefineDet.