Underwater object detection in sonar imagery with detection transformer and Zero-shot neural architecture search
XiaoTong Gu, Shengyu Tang, Yiming Cao, Changdong Yu

TL;DR
This paper introduces NAS-DETR, a novel neural architecture search optimized detection transformer for underwater sonar image object detection, achieving state-of-the-art results with real-time efficiency and enhanced interpretability.
Contribution
It is the first to combine DETR with NAS for sonar image detection, optimizing architecture for low-resolution sonar data.
Findings
Achieves state-of-the-art detection accuracy on benchmark datasets.
Maintains real-time processing with minimal computational overhead.
Enhances interpretability through correlation analysis of key parameters.
Abstract
Underwater object detection using sonar imagery has become a critical and rapidly evolving research domain within marine technology. However, sonar images are characterized by lower resolution and sparser features compared to optical images, which seriously degrades the performance of object detection.To address these challenges, we specifically propose a Detection Transformer (DETR) architecture optimized with a Neural Architecture Search (NAS) approach called NAS-DETR for object detection in sonar images. First, an improved Zero-shot Neural Architecture Search (NAS) method based on the maximum entropy principle is proposed to identify a real-time, high-representational-capacity CNN-Transformer backbone for sonar image detection. This method enables the efficient discovery of high-performance network architectures with low computational and time overhead. Subsequently, the backbone is…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsUnderwater Acoustics Research · Advanced Neural Network Applications · Image Enhancement Techniques
MethodsAttention Is All You Need · Feedforward Network · Linear Layer · Convolution · Multi-Head Attention · Dense Connections · Detection Transformer · Dropout · Layer Normalization · Position-Wise Feed-Forward Layer
