IDEA-Research/DINO
Official implementation of a transformer-based object detection model achieving state-of-the-art performance on COCO benchmarks.
Not currently ranked — collecting fresh signals.
star history
DINO implements an end-to-end object detector based on DETR with improved denoising anchor boxes. It uses a transformer architecture with contrastive training and a DN (denoising) mechanism to achieve leading accuracy on COCO object detection benchmarks. The repository provides training code, evaluation pipelines, and pretrained checkpoints for the detection model.
Frequently asked
- What is IDEA-Research/DINO?
- Official implementation of a transformer-based object detection model achieving state-of-the-art performance on COCO benchmarks.
- Is DINO open source?
- Yes — IDEA-Research/DINO is open source, released under the Apache-2.0 license.
- What language is DINO written in?
- IDEA-Research/DINO is primarily written in Python.
- How popular is DINO?
- IDEA-Research/DINO has 2.8k stars on GitHub.
- Where can I find DINO?
- IDEA-Research/DINO is on GitHub at https://github.com/IDEA-Research/DINO.