Search papers, labs, and topics across Lattice.
The paper introduces Semantic Class Distribution Learning (SCDL), a plug-and-play module designed to mitigate supervision and representation biases in semi-supervised medical image segmentation caused by class imbalance. SCDL employs Class Distribution Bidirectional Alignment (CDBA) to align image embeddings with learnable class proxies and Semantic Anchor Constraints (SAC) to guide these proxies using labeled data. Experiments on Synapse and AMOS datasets show that SCDL significantly improves segmentation performance, especially for minority classes, achieving state-of-the-art results.
Class imbalance in medical image segmentation is no longer a bottleneck: SCDL learns structured class-conditional feature distributions to segment minority structures with state-of-the-art accuracy.
Medical image segmentation is critical for computer-aided diagnosis. However, dense pixel-level annotation is time-consuming and expensive, and medical datasets often exhibit severe class imbalance. Such imbalance causes minority structures to be overwhelmed by dominant classes in feature representations, hindering the learning of discriminative features and making reliable segmentation particularly challenging. To address this, we propose the Semantic Class Distribution Learning (SCDL) framework, a plug-and-play module that mitigates supervision and representation biases by learning structured class-conditional feature distributions. SCDL integrates Class Distribution Bidirectional Alignment (CDBA) to align embeddings with learnable class proxies and leverages Semantic Anchor Constraints (SAC) to guide proxies using labeled data. Experiments on the Synapse and AMOS datasets demonstrate that SCDL significantly improves segmentation performance across both overall and class-level metrics, with particularly strong gains on minority classes, achieving state-of-the-art results. Our code is released at https://github.com/Zyh55555/SCDL.
