Papers
Topics
Authors
Recent
2000 character limit reached

A Fourier Transform Framework for Domain Adaptation (2403.07798v2)

Published 12 Mar 2024 in cs.CV

Abstract: By using unsupervised domain adaptation (UDA), knowledge can be transferred from a label-rich source domain to a target domain that contains relevant information but lacks labels. Many existing UDA algorithms suffer from directly using raw images as input, resulting in models that overly focus on redundant information and exhibit poor generalization capability. To address this issue, we attempt to improve the performance of unsupervised domain adaptation by employing the Fourier method (FTF).Specifically, FTF is inspired by the amplitude of Fourier spectra, which primarily preserves low-level statistical information. In FTF, we effectively incorporate low-level information from the target domain into the source domain by fusing the amplitudes of both domains in the Fourier domain. Additionally, we observe that extracting features from batches of images can eliminate redundant information while retaining class-specific features relevant to the task. Building upon this observation, we apply the Fourier Transform at the data stream level for the first time. To further align multiple sources of data, we introduce the concept of correlation alignment. To evaluate the effectiveness of our FTF method, we conducted evaluations on four benchmark datasets for domain adaptation, including Office-31, Office-Home, ImageCLEF-DA, and Office-Caltech. Our results demonstrate superior performance.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (44)
  1. Gnet: 3d object detection from point cloud with geometry-aware network. In: 2022 IEEE International Conference on Cyborg and Bionic Systems (CBS). 2023, p. 190–195. doi:10.1109/CBS55922.2023.10115327.
  2. Houghlanenet: Lane detection with deep hough transform and dynamic convolution. Computers & Graphics 2023;116:82–92. URL: https://www.sciencedirect.com/science/article/pii/S0097849323001814. doi:https://doi.org/10.1016/j.cag.2023.08.012.
  3. Deeplab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected crfs. IEEE Transactions on Pattern Analysis and Machine Intelligence 2018;40(4):834–848. doi:10.1109/TPAMI.2017.2699184.
  4. Developable mesh segmentation by detecting curve-like features on gauss images. Computers & Graphics 2022;109:42–54. URL: https://www.sciencedirect.com/science/article/pii/S0097849322001819. doi:https://doi.org/10.1016/j.cag.2022.10.003.
  5. Dynamic and multi-source semantic annotation of raw mobility data using geographic and social media data. Pervasive Mob Comput 2021;71:101310. URL: https://api.semanticscholar.org/CorpusID:232328241.
  6. Mixed reality annotation of robotic-assisted surgery videos with real- time tracking and stereo matching. Computers & Graphics 2023;110:125–140. URL: https://www.sciencedirect.com/science/article/pii/S0097849322002291. doi:https://doi.org/10.1016/j.cag.2022.12.006.
  7. A multi-view multi-scale neural network for multi-label ecg classification. IEEE Transactions on Emerging Topics in Computational Intelligence 2023;7(3):648–660. doi:10.1109/TETCI.2023.3235374.
  8. A discrete-time projection neural network for sparse signal reconstruction with application to face recognition. IEEE Transactions on Neural Networks and Learning Systems 2018;30(1):151–162.
  9. Few-shot domain adaptation via mixup optimal transport. IEEE Transactions on Image Processing 2022;31:2518–2528. doi:10.1109/TIP.2022.3157139.
  10. Semi-supervised low-rank semantics grouping for zero-shot learning. IEEE Transactions on Image Processing 2021a;30:2207–2219. doi:10.1109/TIP.2021.3050677.
  11. Domain-adversarial training of neural networks. Journal of Machine Learning Research 2016;17(59):1–35. URL: http://jmlr.org/papers/v17/15-239.html.
  12. Conditional adversarial domain adaptation. In: Neural Information Processing Systems. 2017a,URL: https://api.semanticscholar.org/CorpusID:46784066.
  13. Adversarial graph augmentation to improve graph contrastive learning. In: Neural Information Processing Systems. 2021,URL: https://api.semanticscholar.org/CorpusID:235391029.
  14. A closer look at smoothness in domain adversarial training. In: Chaudhuri, K, Jegelka, S, Song, L, Szepesvari, C, Niu, G, Sabato, S, editors. Proceedings of the 39th International Conference on Machine Learning; vol. 162 of Proceedings of Machine Learning Research. PMLR; 2022, p. 18378–18399. URL: https://proceedings.mlr.press/v162/rangwani22a.html.
  15. Fda: Fourier domain adaptation for semantic segmentation. In: 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 2020, p. 4084–4094. doi:10.1109/CVPR42600.2020.00414.
  16. CyCADA: Cycle-consistent adversarial domain adaptation. In: Dy, J, Krause, A, editors. Proceedings of the 35th International Conference on Machine Learning; vol. 80 of Proceedings of Machine Learning Research. PMLR; 2018, p. 1989–1998. URL: https://proceedings.mlr.press/v80/hoffman18a.html.
  17. Discriminative manifold distribution alignment for domain adaptation. IEEE Transactions on Systems, Man, and Cybernetics: Systems 2023;53(2):1183–1197. doi:10.1109/TSMC.2022.3195239.
  18. Unpaired image-to-image translation using cycle-consistent adversarial networks. In: 2017 IEEE International Conference on Computer Vision (ICCV). 2017, p. 2242–2251. doi:10.1109/ICCV.2017.244.
  19. Visual domain adaptation with manifold embedded distribution alignment. Proceedings of the 26th ACM international conference on Multimedia 2018;URL: https://api.semanticscholar.org/CorpusID:49883347.
  20. Self-training with noisy student improves imagenet classification. In: 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 2020, p. 10684–10695. doi:10.1109/CVPR42600.2020.01070.
  21. Feather-light fourier domain adaptation in magnetic resonance imaging. In: Kamnitsas, K, Koch, L, Islam, M, Xu, Z, Cardoso, J, Dou, Q, et al., editors. Domain Adaptation and Representation Transfer. Cham: Springer Nature Switzerland. ISBN 978-3-031-16852-9; 2022, p. 88–97.
  22. Boosting unsupervised domain adaptation: A fourier approach. Knowledge-Based Systems 2023;264:110325. URL: https://www.sciencedirect.com/science/article/pii/S0950705123000758. doi:https://doi.org/10.1016/j.knosys.2023.110325.
  23. Rda: Robust domain adaptation via fourier adversarial attacking. 2021 IEEE/CVF International Conference on Computer Vision (ICCV) 2021;:8968–8979URL: https://api.semanticscholar.org/CorpusID:235359088.
  24. Minimizing-entropy and fourier consistency network for domain adaptation on optic disc and cup segmentation. IEEE Access 2021b;9:153985–153994. doi:10.1109/ACCESS.2021.3128174.
  25. Source free domain adaptation for medical image segmentation with fourier style mining. Medical Image Analysis 2022;79:102457. URL: https://www.sciencedirect.com/science/article/pii/S1361841522001049. doi:https://doi.org/10.1016/j.media.2022.102457.
  26. Fftw: an adaptive software architecture for the fft. In: Proceedings of the 1998 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP ’98 (Cat. No.98CH36181); vol. 3. 1998, p. 1381–1384 vol.3. doi:10.1109/ICASSP.1998.681704.
  27. ImageCLEF, . Imageclef - the clef cross language image retrieval track; 2014. Accessed 25 June 2023. https://www.imageclef.org/.
  28. Adapting visual category models to new domains. In: Daniilidis, K, Maragos, P, Paragios, N, editors. Computer Vision – ECCV 2010. Berlin, Heidelberg: Springer Berlin Heidelberg. ISBN 978-3-642-15561-1; 2010, p. 213–226.
  29. Geodesic flow kernel for unsupervised domain adaptation. In: 2012 IEEE Conference on Computer Vision and Pattern Recognition. 2012, p. 2066–2073. doi:10.1109/CVPR.2012.6247911.
  30. Gradient-based learning applied to document recognition. Proceedings of the IEEE 1998;86(11):2278–2324. doi:10.1109/5.726791.
  31. Domain adaptation via transfer component analysis. IEEE Transactions on Neural Networks 2011;22(2):199–210. doi:10.1109/TNN.2010.2091281.
  32. Transfer feature learning with joint distribution adaptation. In: Proceedings of the IEEE International Conference on Computer Vision (ICCV). 2013,.
  33. Deep residual learning for image recognition. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). 2016,.
  34. Deep transfer learning with joint adaptation networks. In: Precup, D, Teh, YW, editors. Proceedings of the 34th International Conference on Machine Learning; vol. 70 of Proceedings of Machine Learning Research. PMLR; 2017b, p. 2208–2217. URL: https://proceedings.mlr.press/v70/long17a.html.
  35. Cluster alignment with a teacher for unsupervised domain adaptation. In: 2019 IEEE/CVF International Conference on Computer Vision (ICCV). 2019, p. 9943–9952. doi:10.1109/ICCV.2019.01004.
  36. Deep coral: Correlation alignment for deep domain adaptation. In: ECCV Workshops. 2016,URL: https://api.semanticscholar.org/CorpusID:12453047.
  37. Multi-representation adaptation network for cross-domain image classification. Neural networks : the official journal of the International Neural Network Society 2019;119:214–221. URL: https://api.semanticscholar.org/CorpusID:201700545.
  38. Unsupervised domain adaptation via discriminative manifold propagation. IEEE Transactions on Pattern Analysis and Machine Intelligence 2022;44(3):1653–1669. doi:10.1109/TPAMI.2020.3014218.
  39. Dynamic weighted learning for unsupervised domain adaptation. 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2021;:15237–15246URL: https://api.semanticscholar.org/CorpusID:232352345.
  40. Unsupervised domain adaptation with structural attribute learning networks. Neurocomputing 2020;415:96–105. URL: https://www.sciencedirect.com/science/article/pii/S0925231220311620. doi:https://doi.org/10.1016/j.neucom.2020.07.054.
  41. Transductive learning via improved geodesic sampling. In: British Machine Vision Conference. 2019,URL: https://api.semanticscholar.org/CorpusID:198234065.
  42. Balamurali, M. t-Distributed Stochastic Neighbor Embedding. Cham: Springer International Publishing. ISBN 978-3-030-26050-7; 2020, p. 1–9. URL: https://doi.org/10.1007/978-3-030-26050-7_446-1. doi:10.1007/978-3-030-26050-7_446-1.
  43. Domain adaptation: Learning bounds and algorithms. ArXiv 2009;abs/0902.3430. URL: https://api.semanticscholar.org/CorpusID:6178817.
  44. A theory of learning from different domains. Machine Learning 2010;79:151–175. URL: https://api.semanticscholar.org/CorpusID:8577357.

Summary

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

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

Whiteboard

Video Overview

Open Problems

We haven't generated a list of open problems mentioned in this paper yet.

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

Continue Learning

We haven't generated follow-up questions for this paper yet.

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

Collections

Sign up for free to add this paper to one or more collections.

User Circle Single Streamline Icon: https://streamlinehq.com Sign Up