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

GLC++: Source-Free Universal Domain Adaptation through Global-Local Clustering and Contrastive Affinity Learning (2403.14410v1)

Published 21 Mar 2024 in cs.CV, cs.AI, and cs.LG

Abstract: Deep neural networks often exhibit sub-optimal performance under covariate and category shifts. Source-Free Domain Adaptation (SFDA) presents a promising solution to this dilemma, yet most SFDA approaches are restricted to closed-set scenarios. In this paper, we explore Source-Free Universal Domain Adaptation (SF-UniDA) aiming to accurately classify "known" data belonging to common categories and segregate them from target-private "unknown" data. We propose a novel Global and Local Clustering (GLC) technique, which comprises an adaptive one-vs-all global clustering algorithm to discern between target classes, complemented by a local k-NN clustering strategy to mitigate negative transfer. Despite the effectiveness, the inherent closed-set source architecture leads to uniform treatment of "unknown" data, impeding the identification of distinct "unknown" categories. To address this, we evolve GLC to GLC++, integrating a contrastive affinity learning strategy. We examine the superiority of GLC and GLC++ across multiple benchmarks and category shift scenarios. Remarkably, in the most challenging open-partial-set scenarios, GLC and GLC++ surpass GATE by 16.7% and 18.6% in H-score on VisDA, respectively. GLC++ enhances the novel category clustering accuracy of GLC by 4.3% in open-set scenarios on Office-Home. Furthermore, the introduced contrastive learning strategy not only enhances GLC but also significantly facilitates existing methodologies.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (81)
  1. M. Long, J. Wang, G. Ding, J. Sun, and P. S. Yu, “Transfer feature learning with joint distribution adaptation,” in ICCV, 2013.
  2. Y. Ganin, E. Ustinova, H. Ajakan, P. Germain, H. Larochelle, F. Laviolette, M. Marchand, and V. Lempitsky, “Domain-adversarial training of neural networks,” JMLR, 2016.
  3. J. Hoffman, E. Tzeng, T. Park, J.-Y. Zhu, P. Isola, K. Saenko, A. Efros, and T. Darrell, “Cycada: Cycle-consistent adversarial domain adaptation,” in ICML, 2018.
  4. K. Saito, K. Watanabe, Y. Ushiku, and T. Harada, “Maximum classifier discrepancy for unsupervised domain adaptation,” in CVPR, 2018.
  5. P. Voigt and A. Von dem Bussche, “The eu general data protection regulation (gdpr),” A Practical Guide, 1st Ed., Cham: Springer International Publishing, vol. 10, no. 3152676, pp. 10–5555, 2017.
  6. J. Liang, D. Hu, and J. Feng, “Do we really need to access the source data? source hypothesis transfer for unsupervised domain adaptation,” in ICML, 2020.
  7. S. Yang, Y. Wang, J. van de Weijer, L. Herranz, and S. Jui, “Generalized source-free domain adaptation,” in ICCV, 2021.
  8. S. Qu, G. Chen, J. Zhang, Z. Li, W. He, and D. Tao, “Bmd: A general class-balanced multicentric dynamic prototype strategy for source-free domain adaptation,” in ECCV.   Springer, 2022.
  9. Z. Cao, L. Ma, M. Long, and J. Wang, “Partial adversarial domain adaptation,” in ECCV, 2018.
  10. Z. Cao, M. Long, J. Wang, and M. I. Jordan, “Partial transfer learning with selective adversarial networks,” in CVPR, 2018.
  11. P. Panareda Busto and J. Gall, “Open set domain adaptation,” in ICCV, 2017.
  12. K. Saito, S. Yamamoto, Y. Ushiku, and T. Harada, “Open set domain adaptation by backpropagation,” in ECCV, 2018.
  13. K. You, M. Long, Z. Cao, J. Wang, and M. I. Jordan, “Universal domain adaptation,” in CVPR, 2019.
  14. K. Saito and K. Saenko, “Ovanet: One-vs-all network for universal domain adaptation,” in ICCV, 2021.
  15. K. Saito, D. Kim, S. Sclaroff, and K. Saenko, “Universal domain adaptation through self supervision,” in NeurIPS, 2020.
  16. G. Li, G. Kang, Y. Zhu, Y. Wei, and Y. Yang, “Domain consensus clustering for universal domain adaptation,” in CVPR, 2021.
  17. L. Chen, Y. Lou, J. He, T. Bai, and M. Deng, “Geometric anchor correspondence mining with uncertainty modeling for universal domain adaptation,” in CVPR, 2022.
  18. J. N. Kundu, N. Venkat, R. V. Babu et al., “Universal source-free domain adaptation,” in CVPR, 2020.
  19. J. Liang, D. Hu, J. Feng, and R. He, “Umad: Universal model adaptation under domain and category shift,” arXiv preprint arXiv:2112.08553, 2021.
  20. P. J. Rousseeuw, “Silhouettes: a graphical aid to the interpretation and validation of cluster analysis,” Journal of computational and applied mathematics, vol. 20, pp. 53–65, 1987.
  21. K. Saenko, B. Kulis, M. Fritz, and T. Darrell, “Adapting visual category models to new domains,” in ECCV, 2010.
  22. H. Venkateswara, J. Eusebio, S. Chakraborty, and S. Panchanathan, “Deep hashing network for unsupervised domain adaptation,” in CVPR, 2017.
  23. X. Peng, B. Usman, N. Kaushik, J. Hoffman, D. Wang, and K. Saenko, “Visda: The visual domain adaptation challenge,” arXiv preprint arXiv:1710.06924, 2017.
  24. X. Peng, Q. Bai, X. Xia, Z. Huang, K. Saenko, and B. Wang, “Moment matching for multi-source domain adaptation,” in ICCV, 2019.
  25. S. Qu, T. Zou, F. Röhrbein, C. Lu, G. Chen, D. Tao, and C. Jiang, “Upcycling models under domain and category shift,” in CVPR, 2023.
  26. S. J. Pan and Q. Yang, “A survey on transfer learning,” IEEE TKDE, 2009.
  27. W. Zellinger, T. Grubinger, E. Lughofer, T. Natschläger, and S. Saminger-Platz, “Central moment discrepancy (cmd) for domain-invariant representation learning,” in ICLR, 2017.
  28. N. Courty, R. Flamary, D. Tuia, and A. Rakotomamonjy, “Optimal transport for domain adaptation,” IEEE TPMAI, 2016.
  29. G. Kang, L. Jiang, Y. Yang, and A. G. Hauptmann, “Contrastive adaptation network for unsupervised domain adaptation,” in CVPR, 2019.
  30. M. Ghifary, W. B. Kleijn, M. Zhang, D. Balduzzi, and W. Li, “Deep reconstruction-classification networks for unsupervised domain adaptation,” in ECCV, 2016.
  31. K. Bousmalis, G. Trigeorgis, N. Silberman, D. Krishnan, and D. Erhan, “Domain separation networks,” in NeurIPS, 2016.
  32. Z. Murez, S. Kolouri, D. Kriegman, R. Ramamoorthi, and K. Kim, “Image to image translation for domain adaptation,” in CVPR, 2018.
  33. I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y. Bengio, “Generative adversarial networks,” Communications of the ACM, vol. 63, no. 11, pp. 139–144, 2020.
  34. M. Long, Z. Cao, J. Wang, and M. I. Jordan, “Conditional adversarial domain adaptation,” in NeurIPS, 2018.
  35. L. Chen, H. Chen, Z. Wei, X. Jin, X. Tan, Y. Jin, and E. Chen, “Reusing the task-specific classifier as a discriminator: Discriminator-free adversarial domain adaptation,” in CVPR, 2022.
  36. W. Tranheden, V. Olsson, J. Pinto, and L. Svensson, “Dacs: Domain adaptation via cross-domain mixed sampling,” in WACV, 2021.
  37. L. Hoyer, D. Dai, and L. Van Gool, “Hrda: Context-aware high-resolution domain-adaptive semantic segmentation,” in ECCV, 2022.
  38. L. Chen, Z. Wei, X. Jin, H. Chen, M. Zheng, K. Chen, and Y. Jin, “Deliberated domain bridging for domain adaptive semantic segmentation,” in NeurIPS, 2022.
  39. X. Liu, W. Li, Q. Yang, B. Li, and Y. Yuan, “Towards robust adaptive object detection under noisy annotations,” in CVPR, 2022.
  40. H.-K. Hsu, C.-H. Yao, Y.-H. Tsai, W.-C. Hung, H.-Y. Tseng, M. Singh, and M.-H. Yang, “Progressive domain adaptation for object detection,” in WACV, 2020.
  41. J. Jiang, B. Chen, J. Wang, and M. Long, “Decoupled adaptation for cross-domain object detection,” in ICLR, 2022.
  42. J. Liang, Y. Wang, D. Hu, R. He, and J. Feng, “A balanced and uncertainty-aware approach for partial domain adaptation,” in ECCV, 2020.
  43. Z. Fang, J. Lu, F. Liu, J. Xuan, and G. Zhang, “Open set domain adaptation: Theoretical bound and algorithm,” IEEE TNNLS, vol. 32, no. 10, pp. 4309–4322, 2020.
  44. J. N. Kundu, N. Venkat, A. Revanur, R. V. Babu et al., “Towards inheritable models for open-set domain adaptation,” in CVPR, 2020.
  45. Z. Liu, G. Chen, Z. Li, Y. Kang, S. Qu, and C. Jiang, “Psdc: A prototype-based shared-dummy classifier model for open-set domain adaptation,” IEEE Transactions on Cybernetics, 2022.
  46. S. Yang, Y. Wang, K. Wang, S. Jui, and J. van de Weijer, “One ring to bring them all: Towards open-set recognition under domain shift,” arXiv preprint arXiv: 2206.03600, 2022.
  47. R. Li, Q. Jiao, W. Cao, H.-S. Wong, and S. Wu, “Model adaptation: Unsupervised domain adaptation without source data,” in CVPR, 2020.
  48. J. Liang, D. Hu, Y. Wang, R. He, and J. Feng, “Source data-absent unsupervised domain adaptation through hypothesis transfer and labeling transfer,” IEEE TPAMI, 2021.
  49. S. Yang, Y. Wang, J. van de Weijer, L. Herranz, and S. Jui, “Exploiting the intrinsic neighborhood structure for source-free domain adaptation,” in NeurIPS, 2021.
  50. M. Ye, J. Zhang, J. Ouyang, and D. Yuan, “Source data-free unsupervised domain adaptation for semantic segmentation,” in ACM MM, 2021.
  51. Y. Liu, W. Zhang, and J. Wang, “Source-free domain adaptation for semantic segmentation,” in CVPR, 2021.
  52. A. v. d. Oord, Y. Li, and O. Vinyals, “Representation learning with contrastive predictive coding,” arXiv preprint arXiv:1807.03748, 2018.
  53. Y. Tian, D. Krishnan, and P. Isola, “Contrastive multiview coding,” in Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XI 16.   Springer, 2020, pp. 776–794.
  54. K. He, H. Fan, Y. Wu, S. Xie, and R. Girshick, “Momentum contrast for unsupervised visual representation learning,” in CVPR, 2020.
  55. R. D. Hjelm, A. Fedorov, S. Lavoie-Marchildon, K. Grewal, P. Bachman, A. Trischler, and Y. Bengio, “Learning deep representations by mutual information estimation and maximization,” arXiv preprint arXiv:1808.06670, 2018.
  56. T. Chen, S. Kornblith, M. Norouzi, and G. Hinton, “A simple framework for contrastive learning of visual representations,” in ICML.   PMLR, 2020.
  57. Z. Zhang, W. Chen, H. Cheng, Z. Li, S. Li, L. Lin, and G. Li, “Divide and contrast: Source-free domain adaptation via adaptive contrastive learning,” in NeurIPS, 2022.
  58. Y. Liu, Y. Chen, W. Dai, M. Gou, C.-T. Huang, and H. Xiong, “Source-free domain adaptation with contrastive domain alignment and self-supervised exploration for face anti-spoofing,” in ECCV, 2022.
  59. J. Huang, D. Guan, A. Xiao, and S. Lu, “Model adaptation: Historical contrastive learning for unsupervised domain adaptation without source data,” in NeurIPS, 2021.
  60. Z. Qiu, Y. Zhang, H. Lin, S. Niu, Y. Liu, Q. Du, and M. Tan, “Source-free domain adaptation via avatar prototype generation and adaptation,” in IJCAI, 2021.
  61. K. Han, A. Vedaldi, and A. Zisserman, “Learning to discover novel visual categories via deep transfer clustering,” in ICCV, 2019.
  62. Y.-C. Hsu, Z. Lv, and Z. Kira, “Learning to cluster in order to transfer across domains and tasks,” in ICLR, 2018.
  63. Y.-C. Hsu, Z. Lv, J. Schlosser, P. Odom, and Z. Kira, “Multi-class classification without multi-class labels,” in ICLR, 2019.
  64. B. Zhao and K. Han, “Novel visual category discovery with dual ranking statistics and mutual knowledge distillation,” in NeurIPS, 2021.
  65. S. Vaze, K. Han, A. Vedaldi, and A. Zisserman, “Generalized category discovery,” in CVPR, 2022.
  66. D.-H. Lee et al., “Pseudo-label: The simple and efficient semi-supervised learning method for deep neural networks,” in ICML Worshop, 2013.
  67. Y. Pan, T. Yao, Y. Li, Y. Wang, C.-W. Ngo, and T. Mei, “Transferrable prototypical networks for unsupervised domain adaptation,” in CVPR, 2019.
  68. P. Zhang, B. Zhang, T. Zhang, D. Chen, Y. Wang, and F. Wen, “Prototypical pseudo label denoising and target structure learning for domain adaptive semantic segmentation,” in CVPR, 2021.
  69. J. MacQueen et al., “Some methods for classification and analysis of multivariate observations,” in Proceedings of the fifth Berkeley symposium on mathematical statistics and probability.   Oakland, CA, USA, 1967.
  70. T. Caliński and J. Harabasz, “A dendrite method for cluster analysis,” Communications in Statistics-theory and Methods, vol. 3, no. 1, pp. 1–27, 1974.
  71. D. L. Davies and D. W. Bouldin, “A cluster separation measure,” IEEE TPAMI, 1979.
  72. R. Tibshirani, G. Walther, and T. Hastie, “Estimating the number of clusters in a data set via the gap statistic,” Journal of the Royal Statistical Society: Series B (Statistical Methodology), vol. 63, no. 2, pp. 411–423, 2001.
  73. B. Fu, Z. Cao, M. Long, and J. Wang, “Learning to detect open classes for universal domain adaptation,” in ECCV, 2020.
  74. C. E. Shannon, “A mathematical theory of communication,” The Bell system technical journal, vol. 27, no. 3, pp. 379–423, 1948.
  75. H. W. Kuhn, “The hungarian method for the assignment problem,” Naval research logistics quarterly, vol. 2, no. 1-2, pp. 83–97, 1955.
  76. S. Bucci, M. R. Loghmani, and T. Tommasi, “On the effectiveness of image rotation for open set domain adaptation,” in ECCV, 2020.
  77. H. Liu, Z. Cao, M. Long, J. Wang, and Q. Yang, “Separate to adapt: Open set domain adaptation via progressive separation,” in CVPR, 2019.
  78. Z. Cao, K. You, M. Long, J. Wang, and Q. Yang, “Learning to transfer examples for partial domain adaptation,” in CVPR, 2019.
  79. Y. Zhang, T. Liu, M. Long, and M. Jordan, “Bridging theory and algorithm for domain adaptation,” in ICML, 2019.
  80. K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in CVPR, 2016.
  81. J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li, and L. Fei-Fei, “Imagenet: A large-scale hierarchical image database,” in CVPR, 2009.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (7)
  1. Sanqing Qu (20 papers)
  2. Tianpei Zou (6 papers)
  3. Florian Röhrbein (5 papers)
  4. Cewu Lu (203 papers)
  5. Guang Chen (86 papers)
  6. Dacheng Tao (830 papers)
  7. Changjun Jiang (47 papers)
Citations (1)
View on Semantic Scholar

Summary

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

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

GitHub

  1. GitHub - ispc-lab/GLC-plus (4 stars)