Scene Graph Understanding
Research on unbiased scene graph generation for traffic scene understanding with peer learning strategies.
PerceptionI am currently a researcher at the Shenzhen Institute of Artificial Intelligence and Robotics for Society in Shenzhen. I earned my PhD in Computer and Information Engineering from The Chinese University of Hong Kong, Shenzhen, in 2024. My research focuses on robotics, embodied AI, and machine learning. I was fortunate to work with Professor Yangsheng Xu and Professor Tin Lun Lam during my studies.
One paper accepted by IEEE TITS. Focus: unbiased scene graph generation for traffic scene understanding with peer-learning-based generalization.
One paper accepted by ICASSP. Focus: class-relevance learning for out-of-distribution (OOD) detection in open-world settings.
One paper accepted by IEEE RAL. Focus: self-supervised single-line LiDAR depth completion.
One paper accepted by IEEE TCSVT. Focus: natural image matting with sampling propagation attention and trimap generation.
One paper accepted by IEEE TIM. Focus: structure-aware audio-visual scene classification with graph convolution.
One paper accepted by ICASSP 2023. Focus: blind-spot self-supervised affinity learning for image denoising.
One paper accepted by BMVC 2022. Focus: open-set matting with OOD detection and few-shot adaptation.
One paper accepted by IEEE TIP. Focus: divide-and-co-training for better accuracy-efficiency trade-offs.
Three papers accepted by IROS 2021. Topics include indoor scene recognition, object relation modeling, and BCI-assisted robot navigation.
One paper accepted by ICRA 2021. Focus: long-range hand gesture recognition with an attention-based SSD network.
Interested in the full list? Add your complete publication list here or link out to Google Scholar.
Research on unbiased scene graph generation for traffic scene understanding with peer learning strategies.
PerceptionFeature pyramid attention and graph-based modeling for multi-modal scene recognition.
Multi-modalPerception pipelines for robotics applications, with focus on robustness and real-world deployment.
RoboticsSession co-chair, “Semantic Machine Learning I”.
Finalist Award, Life Long Robotic Vision Challenge of IROS 2019.
Honorable Award, JRC-TEDA 2018 JDX Robotics Challenge (Picking Challenge in Unmanned Supermarket).
Best Student Paper Finalist Award, ICIA 2017.
TA, EIE3080 Microprocessors, CUHKSZ.
TA, CSC1001 Introduction to Computer Science: Programming Methodology, CUHKSZ.
TA, CSC4160 Cloud Computing, CUHKSZ.
TA, EIE3080 Microprocessors, CUHKSZ.
TA, CSC3050 Computer Architecture, CUHKSZ.
TA, EIE3080 Microprocessors, CUHKSZ.
TA, CSC3050 Computer Architecture, CUHKSZ.
Liguang Zhou, Yuhongze Zhou, Tin Lun Lam, Yangsheng Xu, “Context-aware Mixture-of-Experts for Unbiased Scene Graph Generation.”
I am always happy to discuss collaborations and research opportunities. If you are interested in working together, please reach out with a brief introduction and your CV.
Links to datasets and benchmarks can be listed here. Add short descriptions and citations as needed.
Replace this with your dataset name, focus, and access link.
Coming soonBest way to reach me is email.
zhouliguang@cuhk.edu.cn
Shenzhen, China
Google Scholar · GitHub · LinkedIn
Geographic visitor overview based on recent server access logs (auto-updated data source).