Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
110 tokens/sec
GPT-4o
56 tokens/sec
Gemini 2.5 Pro Pro
44 tokens/sec
o3 Pro
6 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

DeLR: Active Learning for Detection with Decoupled Localization and Recognition Query (2312.16931v1)

Published 28 Dec 2023 in cs.CV

Abstract: Active learning has been demonstrated effective to reduce labeling cost, while most progress has been designed for image recognition, there still lacks instance-level active learning for object detection. In this paper, we rethink two key components, i.e., localization and recognition, for object detection, and find that the correctness of them are highly related, therefore, it is not necessary to annotate both boxes and classes if we are given pseudo annotations provided with the trained model. Motivated by this, we propose an efficient query strategy, termed as DeLR, that Decoupling the Localization and Recognition for active query. In this way, we are probably free of class annotations when the localization is correct, and able to assign the labeling budget for more informative samples. There are two main differences in DeLR: 1) Unlike previous methods mostly focus on image-level annotations, where the queried samples are selected and exhausted annotated. In DeLR, the query is based on region-level, and we only annotate the object region that is queried; 2) Instead of directly providing both localization and recognition annotations, we separately query the two components, and thus reduce the recognition budget with the pseudo class labels provided by the model. Experiments on several benchmarks demonstrate its superiority. We hope our proposed query strategy would shed light on researches in active learning in object detection.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (45)
  1. Contextual diversity for active learning. In ECCV, pages 137–153, 2020.
  2. Active learning for deep detection neural networks. In ICCV, pages 3671–3679, 2019.
  3. Hierarchical subquery evaluation for active learning on a graph. In 2014 IEEE Conference on Computer Vision and Pattern Recognition, pages 564–571, 2014.
  4. Weakly supervised deep detection networks. In CVPR, pages 2846–2854, 2016.
  5. A simple framework for contrastive learning of visual representations. arXiv preprint arXiv:2002.05709, 2020.
  6. Active learning for deep object detection via probabilistic modeling. In ICCV, pages 10264–10273, October 2021.
  7. Hierarchical sampling for active learning. International Conference on Machine Learning, 2008.
  8. A general agnostic active learning algorithm. In J. Platt, D. Koller, Y. Singer, and S. Roweis, editors, Advances in Neural Information Processing Systems, volume 20. Curran Associates, Inc., 2008.
  9. D.D.Lewis and W.A.Gale. A sequential algorithm for training text classifiers. SIGIR’94, 1994.
  10. Imagenet: A large-scale hierarchical image database. In Proceedings of the International Conference on Computer Vision, 2009.
  11. An adaptive supervision framework for active learning in object detection. In BMVC, 2019.
  12. The pascal visual object classes (VOC) challenge. IJCV, 88(2):303–338, 2010.
  13. Consistency-based semi-supervised active learning: Towards minimizing labeling cost. CoRR, abs/1910.07153, 2019.
  14. Context aware active learning of activity recognition models. In 2015 IEEE International Conference on Computer Vision (ICCV), pages 4543–4551, 2015.
  15. Momentum contrast for unsupervised visual representation learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 9729–9738, 2020.
  16. Deep residual learning for image recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2016.
  17. One-bit supervision for image classification. In Neural Information Processing Systems, 2020.
  18. Semi-supervised active learning with temporal output discrepancy. In ICCV, 2021.
  19. Multi-class active learning for image classification. In CVPR, pages 2372–2379, 2009.
  20. Localization-aware active learning for object detection. In ACCV, pages 506–522, 2018.
  21. Feature pyramid networks for object detection. In CVPR, pages 936–944, 2017.
  22. Focal loss for dense object detection. IEEE TPAMI, 42(2):318–327, 2020.
  23. Microsoft coco: Common objects in context. In ECCV, 2014.
  24. Active learning using pre-clustering. International Conference on Machine Learning, 2004.
  25. We don’t need no bounding-boxes: Training object class detectors using only human verification. CVPR, pages 854–863, 2016.
  26. Extreme clicking for efficient object annotation. ICCV, pages 4940–4949, 2017.
  27. Training object class detectors with click supervision. CVPR, pages 180–189, 2017.
  28. Faster R-CNN: towards real-time object detection with region proposal networks. In NeurIPS, pages 91–99, 2015.
  29. Ufo22{}^{2}start_FLOATSUPERSCRIPT 2 end_FLOATSUPERSCRIPT: A unified framework towards omni-supervised object detection. In ECCV, 2020.
  30. Margin-based active learning for structured output spaces. In ECML, pages 413–424, 2006.
  31. Active learning for convolutional neural networks: A core-set approach. International Conference on Learning Representation, 2018.
  32. Active learning for convolutional neural networks: A core-set approach. In ICLR, 2018.
  33. Crowdsourcing annotations for visual object detection. In AAAI, 2012.
  34. Humble teachers teach better students for semi-supervised object detection. CVPR, pages 3131–3140, 2021.
  35. C-mil: Continuation multiple instance learning for weakly supervised object detection. In CVPR, pages 2199–2208, 2019.
  36. Min-entropy latent model for weakly supervised object detection. IEEE TPAMI, 41(10):2395–2409, 2019.
  37. Cost-effective active learning for deep image classification. IEEE Transactions on Circuits and Systems for Video Technology, 27(12):2591–2600, Dec 2017.
  38. Omni-detr: Omni-supervised object detection with transformers. CVPR, 2022.
  39. End-to-end semi-supervised object detection with soft teacher. ICCV, pages 3040–3049, 2021.
  40. Interactive self-training with mean teachers for semi-supervised object detection. CVPR, pages 5937–5946, 2021.
  41. Learning loss for active learning. In CVPR, pages 93–102, 2019.
  42. Learning loss for active learning, 2019.
  43. Multiple instance active learning for object detection. In CVPR, pages 5330–5339, June 2021.
  44. One-bit active query with contrastive pairs. In CVPR, 2022.
  45. Deformable detr: Deformable transformers for end-to-end object detection. ICLR, 2021.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (6)
  1. Yuhang Zhang (64 papers)
  2. Yuang Deng (2 papers)
  3. Xiaopeng Zhang (100 papers)
  4. Jie Li (553 papers)
  5. Robert C. Qiu (49 papers)
  6. Qi Tian (314 papers)

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now