mlsanityData to 7 Oct 2026

Decoupled DETR For Few-shot Object Detection

arXiv 2311.11570 ↗First posted 20 Nov 2023cs.CV cs.AIcited by 1

Not on the map. Why?

Abstract (arXiv)

Few-shot object detection (FSOD), an efficient method for addressing the severe data-hungry problem, has been extensively discussed. Current works have significantly advanced the problem in terms of model and data. However, the overall performance of most FSOD methods still does not fulfill the desired accuracy. In this paper we improve the FSOD model to address the severe issue of sample imbalance and weak feature propagation. To alleviate modeling bias from data-sufficient base classes, we examine the effect of decoupling the parameters for classes with sufficient data and classes with few samples in various ways. We design a base-novel categories decoupled DETR (DeDETR) for FSOD. We also explore various types of skip connection between the encoder and decoder for DETR. Besides, we notice that the best outputs could come from the intermediate layer of the decoder instead of the last layer; therefore, we build a unified decoder module that could dynamically fuse the decoder layers as the output feature. We evaluate our model on commonly used datasets such as PASCAL VOC and MSCOCO. Our results indicate that our proposed module could achieve stable improvements of 5% to 10% in both fine-tuning and meta-learning paradigms and has outperformed the highest score in recent works.

Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 45 of its 59 references that are in the map. First posted on arXiv 20 Nov 2023.

Builds on

45 of 59 references.

References · 45

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,604
  3. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 13,958
  4. 2015Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networkscited by 9,617
  5. 2020End-to-End Object Detection with Transformerscited by 7,384
  6. 2020Deformable DETR: Deformable Transformers for End-to-End Object Detectioncited by 3,410
  7. 2013Rich feature hierarchies for accurate object detection and semantic segmentationcited by 2,160
  8. 2020YOLOv4: Optimal Speed and Accuracy of Object Detectioncited by 1,624
  9. 2022DINO: DETR with Improved DeNoising Anchor Boxes for End-to-End Object Detectioncited by 1,306
  10. 2019Large-Scale Long-Tailed Recognition in an Open Worldcited by 704
  11. 2022DAB-DETR: Dynamic Anchor Boxes are Better Queries for DETRcited by 569
  12. 2022DN-DETR: Accelerate DETR Training by Introducing Query DeNoisingcited by 486
33 more
  1. 2020Frustratingly Simple Few-Shot Object Detectioncited by 280
  2. 2018Few-shot Object Detection via Feature Reweightingcited by 265
  3. 2021FSCE: Few-Shot Object Detection via Contrastive Proposal Encodingcited by 168
  4. 2020Overcoming Classifier Imbalance for Long-tail Object Detection with Balanced Group Softmaxcited by 152
  5. 2020Multi-Scale Positive Sample Refinement for Few-Shot Object Detectioncited by 116
  6. 2021DeFRCN: Decoupled Faster R-CNN for Few-Shot Object Detectioncited by 106
  7. 2021Generalized Few-Shot Object Detection without Forgettingcited by 88
  8. 2018RepMet: Representative-based metric learning for classification and one-shot object detectioncited by 85
  9. 2021Beyond Max-Margin: Class Margin Equilibrium for Few-shot Object Detectioncited by 80
  10. 2021Semantic Relation Reasoning for Shot-Stable Few-Shot Object Detectioncited by 79
  11. 2018LSTD: A Low-Shot Transfer Detector for Object Detectioncited by 77
  12. 2022Few-Shot Object Detection with Fully Cross-Transformercited by 64
  13. 2021Meta Faster R-CNN: Towards Accurate Few-Shot Object Detection with Attentive Feature Alignmentcited by 57
  14. 2021Dense Relation Distillation with Context-aware Aggregation for Few-Shot Object Detectioncited by 53
  15. 2021Hallucination Improves Few-Shot Object Detectioncited by 50
  16. 2022Meta-DETR: Image-Level Few-Shot Detection with Inter-Class Correlation Exploitationcited by 48
  17. 2021Universal-Prototype Enhancing for Few-Shot Object Detectioncited by 43
  18. 2021Label, Verify, Correct: A Simple Few Shot Object Detection Methodcited by 43
  19. 2020Restoring Negative Information in Few-Shot Object Detectioncited by 28
  20. 2021Few-Shot Object Detection: A Comprehensive Surveycited by 28
  21. 2022AcroFOD: An Adaptive Method for Cross-domain Few-shot Object Detectioncited by 25
  22. 2023Generating Features with Increased Crop-related Diversity for Few-Shot Object Detectioncited by 21
  23. 2022FS-DETR: Few-Shot DEtection TRansformer with prompting and without re-trainingcited by 20
  24. 2022CD-FSOD: A Benchmark for Cross-domain Few-shot Object Detectioncited by 17
  25. 2022Enhanced Training of Query-Based Object Detection via Selective Query Recollectioncited by 17
  26. 2020Data-Efficient Pretraining via Contrastive Self-Supervisioncited by 14
  27. 2019Dense Scale Network for Crowd Countingcited by 12
  28. 2022Multi-Modal Few-Shot Object Detection with Meta-Learning-Based Cross-Modal Promptingcited by 11
  29. 2022Recent Few-Shot Object Detection Algorithms: A Survey with Performance Comparisoncited by 8
  30. 2023Identification of Novel Classes for Improving Few-Shot Object Detectioncited by 7
  31. 2023Improved Region Proposal Network for Enhanced Few-Shot Object Detectioncited by 5
  32. 2022Few-shot Object Detection with Refined Contrastive Learningcited by 4
  33. 2022Rethinking Skip Connections in Encoder-decoder Networks for Monocular Depth Estimationcited by 1

Built on by

Cited by year

Citing papers per year: 2024: 1.

1 paper.

  1. 2024Cross-domain Multi-modal Few-shot Object Detection via Rich Textcited by 5