Invariant Representation via Decoupling Style and Spurious Features from Images
It addresses a real-world OOD generalization problem for machine learning applications where domain labels are unavailable, representing an incremental advance by combining handling of style and spurious features.
The paper tackles out-of-distribution generalization when both style shifts and spurious features exist without domain labels, proposing IRSS to decouple these factors and achieve improved performance, outperforming traditional methods and solving IRM degradation on benchmark datasets.
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.