Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
[ECCV2024 Oral] Official implementation of the paper "Relation DETR: Exploring Explicit Position Relation Prior for Object Detection"
| Date | Stars |
|---|---|
| 2026-07-30 | 262 |
| 2026-08-09 | 263 |
| 2026-08-17 | 264 |
| 2026-08-26 | 265 |
| 2026-09-10 | 266 |
| 2026-09-14 | 265 |
| 2026-09-30 | 266 |
| 2026-10-16 | 265 |
| 2026-11-04 | 265 |
Today
— stars today
This week
— stars this week
This month
-1 stars this month
Momentum
0.0
growth rate 0.00%/day
**Relation DETR**
===
By [Xiuquan Hou](https://github.com/xiuqhou), [Meiqin Liu](https://scholar.google.com/citations?user=T07OWMkAAAAJ&hl=zh-CN&oi=ao), Senlin Zhang, [Ping Wei](https://scholar.google.com/citations?user=1OQBtdcAAAAJ&hl=zh-CN&oi=ao), [Badong Chen](https://scholar.google.com/citations?user=mq6tPX4AAAAJ&hl=zh-CN&oi=ao), Xuguang Lan.
This repo is the official implementation of [Relation DETR: Exploring Explicit Position Relation Prior for Object Detection](https://arxiv.org/abs/2407.11699v1) accepted to **ECCV2024** (score **5444**, **oral** presentation). [[Arxiv paper link](https://arxiv.org/abs/2407.11699)], [[论文介绍](https://www.bilibili.com/video/BV13m421377m)], [[代码讲解](https://www.bilibili.com/video/BV1uf421B7e7)]
💖 If our [Relation-DETR](https://github.com/xiuqhou/Relation-DETR) or [SA-Det-100k](https://huggingface.co/datasets/xiuqhou/SA-Det-100k/tree/main) dataset is helpful to your researches or projects, please star this repository. Thanks! 🤗
<div align="center">
<img src="images/convergence_curve.png" width=500>
</div>
## TODO
...Want more features? Open a [Feature Request](https://github.com/xiuqhou/Relation-DETR/issues/new?assignees=&labels=enhancement&projects=&template=feature-request-english.yml).
- [x] Support data augmentations from `albumentations`.
- [ ] Support Mosaic and Mixup data augmentation.
- [ ] More detailed docs for the code.
- [ ] Add a instruction about introducing our relation to other models.
- [ ] Support GradCam and feature visualization.
- [x] Upload more pretrained weights and training logs.
- [x] Update visualization code for MC.
- [x] Update Model ZOO.
## Update
- [2024-09-08] Relation-DETR (Focal-Large) checkpoint pretrained on [Object365](https://www.objects365.org/overview.html) is now available [here](https://github.com/xiuqhou/Relation-DETR/releases/download/v1.0.0/relation_detr_focalnet_large_lrf_fl4_800_1333_o365_4e.pth).
- [2024-08-15] We release [config](configs/dino++/dino++_resnet50_800_1333.py) and [checkpoint](https://github.com/xiuqhou/Relation-DETR/releases/download/v1.0.1/dino++_resnet50_800_1333_coco_1x.pth) of DINO++ ([DINO](https://github.com/IDEA-Research/DINO) enhanced by our position relation).
- [2024-08-12] Relation-DETR is selected for **Oral** presentation in ECCV2024!
- [2024-08-11] The pretrained weight for Relation-DETR on SA-Det-100k are available [here](https://github.com/xiuqhou/Relation-DETR/releases/download/v1.0.0/relation_detr_resnet50_800_1333_sa_det_100k_1x.pth)!
- [2024-08-07] Relation-DETR with [FocalNet-large](https://github.com/microsoft/FocalNet/) achieves `63.5AP` on COCO test-dev2017 after fine-tuned for 4 epochs on [Object365](https://www.objects365.org/overview.html), [config](configs/relation_detr/relation_detr_focalnet_large_lrf_fl4_1200_2000.py) and [checkpoint](https://github.com/xiuqhou/Relation-DETR/releases/download/v1.0.0/relation_detr_focalnet_large_lrf_fl4_1200_2000_o365_4e-coco_2x.pth) are available now!
- [2024-07-24] Upload SA-Det-100k dataset, see it in [Hugging Face](https://huggingface.co/datasets/xiuqhou/SA-Det-100k/tree/main) and [Ai Studio](https://aistudio.baidu.com/datasetdetail/286554).
- [2024-07-18] Upload Relation-DETR [training logs](https://github.com/xiuqhou/Relation-DETR/issues/2) for pretrained weights.
- [2024-07-18] We release the code for [Relation-DETR](https://arxiv.org/abs/2407.11699v1), Relation-DETR with Swin-L achieves **58.1 AP**!
- [2024-03-26] Code for [Salience-DETR](https://arxiv.org/abs/2403.16131) is available [here](https://github.com/xiuqhou/Salience-DETR).
- [2024-07-17] We release the checkpoint for Relation-DETR with ResNet-50 and Swin-L backbones, see [Releases v1.0.0](https://github.com/xiuqhou/Relation-DETR/releases/tag/v1.0.0).
- [2024-07-01] Relation-DETR is accepted to ECCV2024. Welcome to your attention!
## SA-Det-100k
[SA-Det-100k](https://github.com/xiuqhou/Relation-DETR) is a large-scale class-agnostic object detection dataset for **Research Purposes onlExcerpt of 16,560 characters
Read on GitHubWould you bet a product on this? Bounded 0–100 and slow moving.
matched fp:6fa10a7433ab94c9, topic:object-detection, desc:object detection, readme:object detection
matched fp:6fa10a7433ab94c9, topic:transformer