CellMamba: Adaptive Mamba for Accurate and Efficient Cell Detection

TL;DR

CellMamba is a lightweight, adaptive one-stage detector for biomedical cell detection that combines novel modules to improve accuracy and efficiency in challenging pathological images.

Contribution

The paper introduces CellMamba, a new adaptive detection framework with novel modules like TMAC and Adaptive Mamba Head for improved cell detection accuracy and efficiency.

Findings

Outperforms CNN, Transformer, and Mamba baselines in accuracy

Reduces model size and inference latency

Effective on public biomedical datasets

Abstract

Cell detection in pathological images presents unique challenges due to densely packed objects, subtle inter-class differences, and severe background clutter. In this paper, we propose CellMamba, a lightweight and accurate one-stage detector tailored for fine-grained biomedical instance detection. Built upon a VSSD backbone, CellMamba integrates CellMamba Blocks, which couple either NC-Mamba or Multi-Head Self-Attention (MSA) with a novel Triple-Mapping Adaptive Coupling (TMAC) module. TMAC enhances spatial discriminability by splitting channels into two parallel branches, equipped with dual idiosyncratic and one consensus attention map, adaptively fused to preserve local sensitivity and global consistency. Furthermore, we design an Adaptive Mamba Head that fuses multi-scale features via learnable weights for robust detection under varying object sizes. Extensive experiments on two public datasets-CoNSeP and CytoDArk0-demonstrate that CellMamba outperforms both CNN-based, Transformer-based, and Mamba-based baselines in accuracy, while significantly reducing model size and inference latency. Our results validate CellMamba as an efficient and effective solution for high-resolution cell detection.