大概总结了一下CVPR2022目标检测领域的文章，并未包括跨域和3D目标检测。 个人总结，难免有疏漏，大家参考一下就好。

CVPR 2022

一、常规目标检测

1. MViTv2: Improved Multiscale Vision Transformers for Classification and Detection

2. Accelerating DETR Convergence via Semantic-Aligned Matching

3. AdaMixer: A Fast-Converging Query-Based Object Detector

4. DESTR: Object Detection With Split Transformer

5. R ( D e t ) 2 R(Det)^2 R(Det)2: Randomized Decision Routing for Object Detection

二、半监督目标检测

6. Dense Learning based Semi-Supervised Object Detection

7. Active Teacher for Semi-Supervised Object Detection

8.Group R-CNN for Weakly Semi-Supervised Object Detection With Points

9.Scale-Equivalent Distillation for Semi-Supervised Object Detection

10.Unbiased Teacher v2: Semi-Supervised Object Detection for Anchor-Free and Anchor-Based Detectors

11.MUM: Mix Image Tiles and UnMix Feature Tiles for Semi-Supervised Object Detection

三、少样本目标检测

12. Label, Verify, Correct: A Simple Few Shot Object Detection Method

13. Semantic-aligned Fusion Transformer for One-shot Object Detection

14. Balanced and Hierarchical Relation Learning for One-shot Object Detection

- 提出一个简单而有效的Ratio-Preserving Loss，以解决正负样本不平衡的问题，从而实现IHR模块的平衡和有效学习。

18. SIOD: Single Instance Annotated Per Category Per Image for Object Detection

创新点 提出了 SIOD 任务，节省了标注成本。 提出挖掘未标记实例的 DMiner 框架，提高对伪标记的容忍度。

16. Few-Shot Object Detection With Fully Cross-Transformer

17. Sylph: A Hypernetwork Framework for Incremental Few-Shot Object Detection

18.Adaptive Hierarchical Representation Learning for Long-Tailed Object Detection

四、知识蒸馏

19. Focal and Global Knowledge Distillation for Detectors

20. Localization Distillation for Dense Object Detection

五、开集目标检测

21. OW-DETR: Open-world Detection Transformer

22. Unknown-Aware Object Detection: Learning What You Don’t Know from Videos in the Wild

23. Expanding Low-Density Latent Regions for Open-Set Object Detection

六、增量目标检测

24. Overcoming Catastrophic Forgetting in Incremental Object Detection via Elastic Response Distillation

七、小目标目标检测

25. QueryDet: Cascaded Sparse Query for Accelerating High-Resolution Small Object Detection

26. Progressive End-to-End Object Detection in Crowded Scenes

八、有向目标检测

27. Oriented RepPoints for Aerial Object Detection

28. OSKDet: Orientation-Sensitive Keypoint Localization for Rotated Object Detection

九、其他

29. Interactron: Embodied Adaptive Object Detection

30. Confidence Propagation Cluster: Unleash Full Potential of Object Detectors