Guidong Yang

Dr. Guidong Yang

orcid

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 January 2026 in IEEE Journal of Oceanic Engineering
Verified
Journal Article
Closed access
JWJunjie Wen
Junjie Wen
Guidong YangGuidong Yang
Guidong Yang
BZBenyun Zhao
Benyun Zhao
LLLei Lei
Lei Lei
ZGZHI GAO
ZHI GAO
BCBen M. Chen
Ben M. Chen
Underwater environments present significant challenges, such as image degradation and domain discrepancies, that severely impact object detection performance. Traditional approaches often use image enhancement as a preprocessing step, but this adds computational overhead, latency, and can even degrade detection accuracy. To address these issues, we propose a novel underwater object detection framework that jointly trains image enhancement within a multitask architecture. This framework employs a progressive training strategy to iteratively improve detection performance through enhancement and introduces a domain-adaptation mechanism to align features across domains at both image and object levels. Experimental results demonstrate that our method achieves state-of-the-art performance across diverse data sets, with real-time detection at 105.93 frames per second and over +15$\%$ mean average precision absolute improvement in unseen environments, underscoring its potential for real-world underwater applications.

24 June 2024
Verified
Conference Paper
Closed access
JWJunjie Wen
Junjie Wen
JCJinqiang Cui
Jinqiang Cui
Guidong YangGuidong Yang
Guidong Yang
BZBenyun Zhao
Benyun Zhao
ZGZhi Gao
Zhi Gao
BCBen M. Chen
Ben M. Chen
Despite the significant achievements made in deep learning for object detection, the challenging conditions of underwater environments severely affect image quality, thereby adversely impacting the effectiveness of detecting underwater objects. Current techniques often employ image enhancement prior to processing to improve the visual quality of underwater images, yet this method incurs extra computational costs and delays, with a risk of diminishing the overall performance of detection. To address these concerns, an efficient framework is presented for underwater object detection, integrating underwater image enhancement with a multi-phase training approach. The framework entails the utilization of both detection and enhancement heads within a unified network backbone, leveraging separate datasets for training purposes. Furthermore, a multi-phase training methodology is introduced to appropriately balance the joint training of detection and enhancement tasks. Extensive experimentation validates that the network proposed in this study attains state-of-the-art performance when compared with other underwater object detectors.