Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

SPMamba-YOLO: An Underwater Object Detection Network Based on Multi-Scale Feature Enhancement and Global Context Modeling

Guanghao Liao, Zhen Liu, Liyuan Cao, Yonghui Yang, Qi Li

TL;DR

A Spatial Pyramid Pooling Enhanced Layer Aggregation Network (SPPELAN) module is introduced to strengthen multi-scale feature aggregation and expand the receptive field, while a Pyramid Split Attention (PSA) mechanism enhances feature discrimination by emphasizing informative regions and suppressing background interference.

Abstract

Underwater object detection is a critical yet challenging research problem owing to severe light attenuation, color distortion, background clutter, and the small scale of underwater targets. To address these challenges, we propose SPMamba-YOLO, a novel underwater object detection network that integrates multi-scale feature enhancement with global context modeling. Specifically, a Spatial Pyramid Pooling Enhanced Layer Aggregation Network (SPPELAN) module is introduced to strengthen multi-scale feature aggregation and expand the receptive field, while a Pyramid Split Attention (PSA) mechanism enhances feature discrimination by emphasizing informative regions and suppressing background interference. In addition, a Mamba-based state space modeling module is incorporated to efficiently capture long-range dependencies and global contextual information, thereby improving detection robustness in complex underwater environments. Extensive experiments on the URPC2022 dataset demonstrate that SPMamba-YOLO outperforms the YOLOv8n baseline by more than 4.9\% in mAP@0.5, particularly for small and densely distributed underwater objects, while maintaining a favorable balance between detection accuracy and computational cost.

SPMamba-YOLO: An Underwater Object Detection Network Based on Multi-Scale Feature Enhancement and Global Context Modeling

TL;DR

A Spatial Pyramid Pooling Enhanced Layer Aggregation Network (SPPELAN) module is introduced to strengthen multi-scale feature aggregation and expand the receptive field, while a Pyramid Split Attention (PSA) mechanism enhances feature discrimination by emphasizing informative regions and suppressing background interference.

Abstract

Underwater object detection is a critical yet challenging research problem owing to severe light attenuation, color distortion, background clutter, and the small scale of underwater targets. To address these challenges, we propose SPMamba-YOLO, a novel underwater object detection network that integrates multi-scale feature enhancement with global context modeling. Specifically, a Spatial Pyramid Pooling Enhanced Layer Aggregation Network (SPPELAN) module is introduced to strengthen multi-scale feature aggregation and expand the receptive field, while a Pyramid Split Attention (PSA) mechanism enhances feature discrimination by emphasizing informative regions and suppressing background interference. In addition, a Mamba-based state space modeling module is incorporated to efficiently capture long-range dependencies and global contextual information, thereby improving detection robustness in complex underwater environments. Extensive experiments on the URPC2022 dataset demonstrate that SPMamba-YOLO outperforms the YOLOv8n baseline by more than 4.9\% in mAP@0.5, particularly for small and densely distributed underwater objects, while maintaining a favorable balance between detection accuracy and computational cost.
Paper Structure (18 sections, 18 equations, 10 figures, 6 tables)

This paper contains 18 sections, 18 equations, 10 figures, 6 tables.

Table of Contents

  1. Introduction
  2. Related work
  3. Underwater object detection
  4. Small object detection
  5. Method
  6. State space models
  7. Overall Architecture and SSM-based Feature Extraction
  8. PSA attention module
  9. SPPELAN Module
  10. Experiments
  11. Datasets
  12. Implementation details
  13. Evaluation metrics
  14. Ablation experiments
  15. Model visualization
  16. ...and 3 more sections

Figures (10)

  • Figure 1: Underwater robot target detection challenges
  • Figure 2: The structure of SPMamba-YOLO
  • Figure 3: The structure of Simple Stem
  • Figure 4: The structure of Vision Clue Merge
  • Figure 5: The structure of the ODSSBlock and its submodules.(a) Overall architecture of ODSSBlock.(b) LS Block.(c) RG Block.(d) SS2D structure.(e) Scan process in SS2D.
  • ...and 5 more figures