awesome-anchor-free-object-detection

for anyone who wants to do research about anchor free object detection.

If you find the awesome paper/code/dataset or have some suggestions, please contact xin.zhang2018@nlpr.ia.ac.cn, nuo.xu@nlpr.ia.ac.cn and xswang@wayne.edu. Thanks for your valuable contribution to the research community 😃

- Recent papers (from 2015)

Statistics: 🔥 code is available & stars >= 100

2020

[arXiv] OneNet: End-to-End One-Stage Object Detection by Classification Cost.[pytorch]:fire:
[arXiv] End-to-End Object Detection with Fully Convolutional Network
[arXiv] Sparse R-CNN: End-to-End Object Detection with Learnable Proposals.[pytorch]:fire:
[arXiv] End-to-End Object Detection with Transformers.[pytorch]:fire:
[arXiv] AutoAssign: Differentiable Label Assignment for Dense Object Detection.
[arXiv] RepPoints V2: Verification Meets Regression for Object Detection. [pytorch]:fire:
[ECCV] Corner Proposal Network for Anchor-free, Two-stage Object Detection. [Available soon]
[ECCV] HoughNet: Integrating near and long-range evidence for bottom-up object detection. [pytorch]
[CVPR] Bridging the Gap Between Anchor-based and Anchor-free Detection via Adaptive Training Sample Selection. [pytorch]:fire:
[CVPR] Soft Anchor-Point Object Detection. [Keras]
[CVPR] CentripetalNet: Pursuing High-quality Keypoint Pairs for Object Detection. [pytorch]:fire:
[arXiv] SaccadeNet: A Fast and Accurate Object Detector. [pytorch]
[arXiv] Localization Uncertainty Estimation for Anchor-Free Object Detection.
[ECCV] Dense RepPoints: Representing Visual Objects with Dense Point Sets. [pytorch]
[ECCV] BorderDet: Border Feature for Dense Object Detection. [pytorch]:fire:
[arXiv] Generalized Focal Loss: Learning Qualified and Distributed Bounding Boxes for Dense Object Detection. [pytorch]:fire:

[ICCV] RepPoints: Point Set Representation for Object Detection. [pytorch]:fire:
[arXiv] Segmentation is All You Need.
[arXiv] FCOS: Fully Convolutional One-Stage Object Detection. [pytorch]:fire:
[arXiv] CenterNet: Keypoint Triplets for Object Detection. [pytorch]:fire:
[arXiv] Objects as Points. [pytorch]:fire:
[arXiv] FoveaBox: Beyond Anchor-based Object Detector. [pytorch]:fire:
[CVPR] Feature Selective Anchor-Free Module for Single-Shot Object Detection. [pytorch]:fire:
[arXiv] ExtremeNet: Bottom-up Object Detection by Grouping Extreme and Center Points. [pytorch]:fire:

[ECCV] CornerNet: Detecting Objects as Paired Keypoints. [pytorch]:fire:
[arXiv] An Anchor-Free Region Proposal Network for Faster R-CNN based Text Detection Approaches.

[CVPR] You Only Look Once: Unified, Real-Time Object Detection. [tensorflow] [darknet]:fire:
[acm multimedia] UnitBox: An Advanced Object Detection Network. [tensorflow]

[arXiv] DenseBox: Unifying Landmark Localization with End to End Object Detection. [caffe]