mlsanityData to 13 Oct 2026

RPCL: A Framework for Improving Cross-Domain Detection with Auxiliary Tasks

arXiv 2104.08689 ↗First posted 18 Apr 2021cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Cross-Domain Detection (XDD) aims to train an object detector using labeled image from a source domain but have good performance in the target domain with only unlabeled images. Existing approaches achieve this either by aligning the feature maps or the region proposals from the two domains, or by transferring the style of source images to that of target image. Contrasted with prior work, this paper provides a complementary solution to align domains by learning the same auxiliary tasks in both domains simultaneously. These auxiliary tasks push image from both domains towards shared spaces, which bridges the domain gap. Specifically, this paper proposes Rotation Prediction and Consistency Learning (PRCL), a framework complementing existing XDD methods for domain alignment by leveraging the two auxiliary tasks. The first one encourages the model to extract region proposals from foreground regions by rotating an image and predicting the rotation angle from the extracted region proposals. The second task encourages the model to be robust to changes in the image space by optimizing the model to make consistent class predictions for region proposals regardless of image perturbations. Experiments show the detection performance can be consistently and significantly enhanced by applying the two proposed tasks to existing XDD methods.

Key facts. No citing papers in the map yet (data through 13 Oct 2026). Builds on 49 of its 57 references that are in the map. First posted on arXiv 18 Apr 2021.

Builds on

49 of 57 references.

References · 49

  1. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,321
  2. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 12,009
  3. 2015Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networkscited by 9,625
  4. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,492
  5. 2017Mask R-CNNcited by 7,266
  6. 2017mixup: Beyond Empirical Risk Minimizationcited by 5,170
  7. 2015You Only Look Once: Unified, Real-Time Object Detectioncited by 4,840
  8. 2016The Cityscapes Dataset for Semantic Urban Scene Understandingcited by 4,435
  9. 2015SSD: Single Shot MultiBox Detectorcited by 3,128
  10. 2015Fast R-CNNcited by 2,933
  11. 2018YOLOv3: An Incremental Improvementcited by 2,655
  12. 2013Rich feature hierarchies for accurate object detection and semantic segmentationcited by 2,162
37 more
  1. 2020FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidencecited by 2,082
  2. 2018Unsupervised Representation Learning by Predicting Image Rotationscited by 1,590
  3. 2016YOLO9000: Better, Faster, Strongercited by 1,430
  4. 2016Context Encoders: Feature Learning by Inpaintingcited by 1,427
  5. 2019MixMatch: A Holistic Approach to Semi-Supervised Learningcited by 1,391
  6. 2016Unsupervised Learning of Visual Representations by Solving Jigsaw Puzzlescited by 1,381
  7. 2016Colorful Image Colorizationcited by 1,114
  8. 2015Unsupervised Visual Representation Learning by Context Predictioncited by 1,064
  9. 2019Unsupervised Data Augmentation for Consistency Trainingcited by 1,058
  10. 2016Temporal Ensembling for Semi-Supervised Learningcited by 1,055
  11. 2017Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learningcited by 919
  12. 2018Domain Adaptive Faster R-CNN for Object Detection in the Wildcited by 442
  13. 2016Regularization With Stochastic Transformations and Perturbations for Deep Semi-Supervised Learningcited by 441
  14. 2017Semantic Foggy Scene Understanding with Synthetic Datacited by 401
  15. 2014Discriminative Unsupervised Feature Learning with Exemplar Convolutional Neural Networkscited by 299
  16. 2018Strong-Weak Distribution Alignment for Adaptive Object Detectioncited by 269
  17. 2019Revisiting Self-Supervised Visual Representation Learningcited by 251
  18. 2018Cross-Domain Weakly-Supervised Object Detection through Progressive Domain Adaptationcited by 218
  19. 2016Split-Brain Autoencoders: Unsupervised Learning by Cross-Channel Predictioncited by 217
  20. 2017Colorization as a Proxy Task for Visual Understandingcited by 175
  21. 2020Universal Domain Adaptation through Self Supervisioncited by 172
  22. 2019Population Based Augmentation: Efficient Learning of Augmentation Policy Schedulescited by 153
  23. 2020Exploring Categorical Regularization for Domain Adaptive Object Detectioncited by 130
  24. 2020Harmonizing Transferability and Discriminability for Adapting Object Detectorscited by 125
  25. 2019Diversify and Match: A Domain Adaptive Representation Learning Paradigm for Object Detectioncited by 122
  26. 2019Unsupervised Domain Adaptation through Self-Supervisioncited by 116
  27. 2019Multi-adversarial Faster-RCNN for Unrestricted Object Detectioncited by 111
  28. 2019A Robust Learning Approach to Domain Adaptive Object Detectioncited by 109
  29. 2020Every Pixel Matters: Center-aware Feature Alignment for Domain Adaptive Object Detectorcited by 106
  30. 2019Self-Training and Adversarial Background Regularization for Unsupervised Domain Adaptive One-Stage Object Detectioncited by 83
  31. 2019When Does Self-supervision Improve Few-shot Learning?cited by 78
  32. 2018Semi-Supervised Sequence Modeling with Cross-View Trainingcited by 75
  33. 2019Automatic adaptation of object detectors to new domains using self-trainingcited by 66
  34. 2019Self-supervised Domain Adaptation for Computer Vision Taskscited by 58
  35. 2020Domain Adaptive Object Detection via Asymmetric Tri-way Faster-RCNNcited by 55
  36. 2020Cross-Domain Document Object Detection: Benchmark Suite and Methodcited by 24
  37. 2020Self-Supervised Domain Adaptation with Consistency Trainingcited by 6

Built on by

Not cited yet.