Cin-Seg: Causal Invariance for Tag-Supervised Segmentation on Medical Images
Zhang Chen, Zhiqiang Tian, Jihua Zhu, Shaoyi Du, Qindong Sun · IEEE Transactions on Instrumentation and Measurement · 2024
Weakly supervised semantic segmentation (WSSS) is the method that learning a segmentation model with only weak labels, e.g., image-level labels. For WSSS methods, the segmentation result cannot be directly learned because of the lack of pixel-level annotation. Existing methods use category-highly-correlated region to approximate pixel-level segmentation result of target object. However, the category-highly-correlated region has no causal relation with segmentation result. In this paper, we propose a causal invariance-based method to learn the intrinsic attribute that is causally related to the segmentation result of target object. We model WSSS task as a causal problem from the perspective of causal invariance based on the fact that the intrinsic attribute of target object is invariant. According to the causal modeling, it is feasible to learn intrinsic attributes of target object without supervision of pixel-level annotation. We propose a causal invariant transformation (CIT) strategy to explicitly supervise the model to learn causal feature of segmentation by using a multi-branch architecture. The CIT constrains the prediction to follow the intrinsic attribute of target object, which improves the accuracy of segmentation result with only tag supervision. Experimental results on three medical data sets show that the proposed method outperforms state-of-the-art WSSS methods. The code is available at https://github.com/cz-xjtu/Cin-Seg.