Enhanced Water Surface Object Detection with Dynamic Task-Aligned Sample Assignment and Attention Mechanisms

TL;DR

This paper introduces a new real-time detection system for objects on water surfaces, improving accuracy with attention mechanisms and dynamic sample assignment.

Contribution

The novel integration of CBAM, self-attention, and dynamic sample assignment improves detection accuracy on water surfaces.

Findings

The model achieves 47.1% mAP on the water surface object dataset, a 1.7% improvement over YOLOv8.

Dynamic sample assignment improves AP0.5 by 1.0%, and FFN refines boundaries with 0.8% AP0.75 improvement.

Ablation studies confirm the approach's versatility for other detection frameworks.

Abstract

The detection of objects on water surfaces is a pivotal technology for the perceptual systems of unmanned surface vehicles (USVs). This paper proposes a novel real-time target detection system designed to address the challenges posed by indistinct bottom boundaries and foggy imagery. Our method enhances the YOLOv8s model by incorporating the convolutional block attention module (CBAM) and a self-attention mechanism, examining their impact at various integration points. A dynamic sample assignment strategy was introduced to enhance the precision of our model and accelerate its convergence. To address the challenge of delineating bottom boundaries with clarity, our model employs a two-strategy approach: a threshold filter and a feedforward neural network (FFN) that provides targeted guidance for refining these boundaries. Our model demonstrated exceptional performance, achieving a mean average precision (mAP) of 47.1% on the water surface object dataset, which represents a 1.7% increase over the baseline YOLOv8 model. The dynamic sample assignment strategy contributes a 1.0% improvement on average precision at the intersection over union (IoU) threshold of 0.5 (AP0.5), while the FFN strategy fine-tunes the bottom boundaries and achieves an additional 0.8% improvement in average precision at IoU threshold of 0.75 (AP0.75). Furthermore, ablation studies have validated the versatility of our approach, confirming its potential for integration into various detection frameworks.

Figures