Enabling Real-Time Colonoscopic Polyp Segmentation on Commodity CPUs via Ultra-Lightweight Architecture

TL;DR

This paper introduces UltraSeg, a lightweight CPU-native model for real-time colonoscopic polyp segmentation, achieving high accuracy and speed without GPU reliance, suitable for resource-limited clinical settings.

Contribution

The work presents UltraSeg, an ultra-lightweight, CPU-efficient segmentation architecture that outperforms existing models in speed and accuracy while maintaining minimal parameters.

Findings

UltraSeg-130K achieves Dice > 0.8 on multiple datasets.

Real-time performance exceeds 50 FPS on a single CPU core.

Outperforms all sub-0.3M competitors and approaches heavyweight models.

Abstract

Real-time polyp segmentation is essential for early colorectal cancer detection, yet clinical deployment remains blocked by GPU dependency. We introduce the UltraSeg family, a set of CPU-native segmentation models operating below 0.3M parameters. UltraSeg-108K (0.108M) establishes the extreme-compression frontier, while UltraSeg-130K (0.130M) integrates cross-layer lightweight fusion for enhanced multi-center generalization. The architecture replaces parameter-heavy components with grouped multi-rate dilated convolutions and attention-gated cross-layer fusion, achieving real-time throughput on a single CPU core (exceeding 50 FPS at 256*256 and 30 FPS at 352*352) without sacrificing clinical-grade accuracy. Evaluated on seven public datasets, UltraSeg-130K attains Dice scores exceeding 0.8 at both resolutions, substantially outperforming all existing sub-0.3M competitors. Notably, it approaches or exceeds UNet-Medium (7.76M parameters) on zero-shot external validations while using only 1.7% of its parameters, establishing the first strong baseline for CPU-native real-time polyp segmentation. When scaled to 4.38M parameters, UltraSeg achieves accuracy competitive with heavyweight state-of-the-art models while maintaining an order-of-magnitude parameter advantage, demonstrating that the proposed design principles yield intrinsic representational gains across the entire efficiency spectrum. By delivering the first clinically deployable, CPU-native real-time solution, this work provides an immediately usable tool for resource-limited settings and a reproducible blueprint for real-time medical AI beyond endoscopy. Source code is publicly available.