DLRMamba: Distilling Low-Rank Mamba for Edge Multispectral Fusion Object Detection

TL;DR

This paper introduces Low-Rank SS2D, a model that reduces parameter redundancy in state space models for multispectral object detection, enabling efficient deployment on edge devices while maintaining high accuracy.

Contribution

The paper proposes Low-Rank SS2D with matrix factorization and a Structure-Aware Distillation strategy to improve efficiency and robustness in multispectral object detection on resource-limited hardware.

Findings

Achieves superior efficiency-accuracy trade-off on benchmark datasets

Reduces computational complexity and memory footprint significantly

Outperforms existing lightweight architectures on edge platforms like Raspberry Pi 5

Abstract

Multispectral fusion object detection is a critical task for edge-based maritime surveillance and remote sensing, demanding both high inference efficiency and robust feature representation for high-resolution inputs. However, current State Space Models (SSMs) like Mamba suffer from significant parameter redundancy in their standard 2D Selective Scan (SS2D) blocks, which hinders deployment on resource-constrained hardware and leads to the loss of fine-grained structural information during conventional compression. To address these challenges, we propose the Low-Rank Two-Dimensional Selective Structured State Space Model (Low-Rank SS2D), which reformulates state transitions via matrix factorization to exploit intrinsic feature sparsity. Furthermore, we introduce a Structure-Aware Distillation strategy that aligns the internal latent state dynamics of the student with a full-rank teacher model to compensate for potential representation degradation. This approach substantially reduces computational complexity and memory footprint while preserving the high-fidelity spatial modeling required for object recognition. Extensive experiments on five benchmark datasets and real-world edge platforms, such as Raspberry Pi 5, demonstrate that our method achieves a superior efficiency-accuracy trade-off, significantly outperforming existing lightweight architectures in practical deployment scenarios.