Diffusion-Guided Mask-Consistent Paired Mixing for Endoscopic Image Segmentation

TL;DR

This paper introduces a novel diffusion-guided paired mixing method for endoscopic image segmentation that enhances robustness and diversity while maintaining accurate masks, achieving state-of-the-art results across multiple datasets.

Contribution

It proposes a paired diffusion-guided paradigm with mask-consistent mixing and adaptive re-anchoring to improve segmentation robustness and generalization.

Findings

Achieves state-of-the-art segmentation performance on multiple datasets.

Enlarges diversity of training samples without losing pixel-level semantics.

Provides robust and generalizable endoscopic segmentation results.

Abstract

Augmentation for dense prediction typically relies on either sample mixing or generative synthesis. Mixing improves robustness but misaligned masks yield soft label ambiguity. Diffusion synthesis increases apparent diversity but, when trained as common samples, overlooks the structural benefit of mask conditioning and introduces synthetic-real domain shift. We propose a paired, diffusion-guided paradigm that fuses the strengths of both. For each real image, a synthetic counterpart is generated under the same mask and the pair is used as a controllable input for Mask-Consistent Paired Mixing (MCPMix), which mixes only image appearance while supervision always uses the original hard mask. This produces a continuous family of intermediate samples that smoothly bridges synthetic and real appearances under shared geometry, enlarging diversity without compromising pixel-level semantics. To keep learning aligned with real data, Real-Anchored Learnable Annealing (RLA) adaptively adjusts the mixing strength and the loss weight of mixed samples over training, gradually re-anchoring optimization to real data and mitigating distributional bias. Across Kvasir-SEG, PICCOLO, CVC-ClinicDB, a private NPC-LES cohort, and ISIC 2017, the approach achieves state-of-the-art segmentation performance and consistent gains over baselines. The results show that combining label-preserving mixing with diffusion-driven diversity, together with adaptive re-anchoring, yields robust and generalizable endoscopic segmentation.