Lightweight Polyp Segmentation via a Gain-Aware Prediction-Space Recursive Controller
2026-07-03 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors want to improve lightweight polyp segmentation methods without relying on complex or heavy model parts. They created a small correction system that works directly on the model’s output predictions, making small improvements in a few steps. This system is easy to use on simple devices and improves accuracy consistently across tests. Their approach also works well on other related datasets without adding much extra cost. The authors tested it with various lightweight models and found it to be effective and efficient.
polyp segmentationlightweight modelsrecursive correctionbackbone logitssegmentation accuracyDice scoreIoU (Intersection over Union)parameter efficiencymodel deploymentKvasir-SEG dataset
Authors
Jiachi Zhang, Zhuoyu Wu, Quanjun Wang, Wenhui Ou, Wenqi Fang
Abstract
While lightweight polyp segmentation is highly desirable for low-cost deployment, reported performance gains often stem from upgraded backbone encoders, complex decoders, or heavy refinement branches. Consequently, it remains difficult to isolate whether a lightweight correction mechanism is inherently effective on its own. We address this limitation by formulating refinement as a prediction-space recursive correction task, introducing a recursive controller that operates directly on backbone logits. Under a fixed recursion budget, this controller aggregates discrepancy and uncertainty evidence, updates a compact state tracking recent correction utility, and applies additive residual logit corrections. By design, this correction path remains small, host-portable, and deployment-explicit. Utilizing a unified Kvasir-trained protocol, we evaluate our approach across seven lightweight backbones on Kvasir-SEG and three transfer datasets, measuring segmentation accuracy (Dice/IoU) alongside deployment efficiency (parameters, GMACs, and peak memory). The controller yields consistent improvements in the source domain, achieves competitive performance against both training-side baselines and heavier structural refiners on representative hosts, and delivers selective transfer gains with minimal static overhead. Code is available at https://github.com/tyui99/Gain-Aware-Prediction-Space-Recursive-Controller.