Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
184 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
45 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Invariant Representation via Decoupling Style and Spurious Features from Images (2312.06226v2)

Published 11 Dec 2023 in cs.CV and cs.AI

Abstract: This paper considers the out-of-distribution (OOD) generalization problem under the setting that both style distribution shift and spurious features exist and domain labels are missing. This setting frequently arises in real-world applications and is underlooked because previous approaches mainly handle either of these two factors. The critical challenge is decoupling style and spurious features in the absence of domain labels. To address this challenge, we first propose a structural causal model (SCM) for the image generation process, which captures both style distribution shift and spurious features. The proposed SCM enables us to design a new framework called IRSS, which can gradually separate style distribution and spurious features from images by introducing adversarial neural networks and multi-environment optimization, thus achieving OOD generalization. Moreover, it does not require additional supervision (e.g., domain labels) other than the images and their corresponding labels. Experiments on benchmark datasets demonstrate that IRSS outperforms traditional OOD methods and solves the problem of Invariant risk minimization (IRM) degradation, enabling the extraction of invariant features under distribution shift.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (32)
  1. Invariant risk minimization. arXiv preprint arXiv:1907.02893.
  2. Decaug: Out-of-distribution generalization via decomposed feature representation and semantic augmentation. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 35, 6705–6713.
  3. Domain generalization by solving jigsaw puzzles. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2229–2238.
  4. Invariant rationalization. In International Conference on Machine Learning, 1448–1458. PMLR.
  5. Grad-cam++: Generalized gradient-based visual explanations for deep convolutional networks. In 2018 IEEE winter conference on applications of computer vision (WACV), 839–847. IEEE.
  6. Pareto invariant risk minimization. arXiv preprint arXiv:2206.07766.
  7. Domain generalization with domain-specific aggregation modules. In Pattern Recognition: 40th German Conference, GCPR 2018, Stuttgart, Germany, October 9-12, 2018, Proceedings 40, 187–198. Springer.
  8. Adversarially adaptive normalization for single domain generalization. In Proceedings of the IEEE/CVF conference on Computer Vision and Pattern Recognition, 8208–8217.
  9. Unbiased metric learning: On the utilization of multiple datasets and web images for softening bias. In Proceedings of the IEEE International Conference on Computer Vision, 1657–1664.
  10. Unsupervised domain adaptation by backpropagation. In International conference on machine learning, 1180–1189. PMLR.
  11. Domain-adversarial training of neural networks. The journal of machine learning research, 17(1): 2096–2030.
  12. In search of lost domain generalization. arXiv preprint arXiv:2007.01434.
  13. Towards non-iid image classification: A dataset and baselines. Pattern Recognition, 110: 107383.
  14. Segment anything. arXiv preprint arXiv:2304.02643.
  15. Out-of-distribution generalization via risk extrapolation (rex). In International Conference on Machine Learning, 5815–5826. PMLR.
  16. Invariant information bottleneck for domain generalization. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 36, 7399–7407.
  17. Deeper, broader and artier domain generalization. In Proceedings of the IEEE international conference on computer vision, 5542–5550.
  18. Demystifying neural style transfer. arXiv preprint arXiv:1701.01036.
  19. Bayesian invariant risk minimization. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 16021–16030.
  20. An empirical study of invariant risk minimization on deep models. In ICML 2021 Workshop on Uncertainty and Robustness in Deep Learning, volume 1, 7.
  21. Learning transferable features with deep adaptation networks. In International conference on machine learning, 97–105. PMLR.
  22. Towards recognizing unseen categories in unseen domains. In European Conference on Computer Vision, 466–483. Springer.
  23. Best sources forward: domain generalization through source-specific nets. In 2018 25th IEEE international conference on image processing (ICIP), 1353–1357. IEEE.
  24. Domain generalization using a mixture of multiple latent domains. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 34, 11749–11756.
  25. Causal inference by using invariant prediction: identification and confidence intervals. Journal of the Royal Statistical Society Series B: Statistical Methodology, 78(5): 947–1012.
  26. The risks of invariant risk minimization. arXiv preprint arXiv:2010.05761.
  27. Unbiased look at dataset bias. In CVPR 2011, 1521–1528. IEEE.
  28. Adversarial discriminative domain adaptation. In Proceedings of the IEEE conference on computer vision and pattern recognition, 7167–7176.
  29. Deep hashing network for unsupervised domain adaptation. In Proceedings of the IEEE conference on computer vision and pattern recognition, 5018–5027.
  30. Deep Stable Learning for Out-Of-Distribution Generalization. arXiv:2104.07876.
  31. Deep domain-adversarial image generation for domain generalisation. In Proceedings of the AAAI conference on artificial intelligence, volume 34, 13025–13032.
  32. Sparse invariant risk minimization. In International Conference on Machine Learning, 27222–27244. PMLR.

Summary

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

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