SAM-DCE: Addressing Token Uniformity and Semantic Over-Smoothing in Medical Segmentation
This work improves medical image segmentation by adapting a general-purpose model to handle domain-specific challenges like anatomical variability, though it appears incremental as it builds on existing prompt-free SAM adaptations.
The paper tackles the problem of adapting the Segment Anything Model (SAM) to medical imaging by addressing token uniformity and semantic over-smoothing issues, resulting in enhanced inter-class separability and fine-grained mask decoding validated on diverse medical benchmarks.
The Segment Anything Model (SAM) demonstrates impressive zero-shot segmentation ability on natural images but encounters difficulties in medical imaging due to domain shifts, anatomical variability, and its reliance on user-provided prompts. Recent prompt-free adaptations alleviate the need for expert intervention, yet still suffer from limited robustness and adaptability, often overlooking the issues of semantic over-smoothing and token uniformity. We propose SAM-DCE, which balances local discrimination and global semantics while mitigating token uniformity, enhancing inter-class separability, and enriching mask decoding with fine-grained, consistent representations. Extensive experiments on diverse medical benchmarks validate its effectiveness.