Embodied AI · VLA systems · Image enhancement

Building efficient visual intelligence for real-world interaction.

I am a Ph.D. student at The University of Sydney, advised by Prof. Chang Xu.

My research focuses on embodied AI, especially the architecture, reasoning, and acceleration of Vision-Language-Action (VLA) and World Action Model (WAM) systems. I also work on low-light image enhancement and machine learning.

01 · UPDATES

News

May 2026

See What Matters is accepted by ICML 2026.

June 2025

FusionNet wins 1st place in the NTIRE 2025 Low-Light Enhancement challenge.

June 2025

HVI is accepted by CVPR 2025 and released as an open-source implementation.

June 2025

Flow-Guided HDR Restoration wins 2nd place in the NTIRE 2025 Burst HDR Restoration challenge.

June 2024

DiffLight receives an oral presentation and 4th place in the NTIRE 2024 Low-Light Enhancement challenge.

02 · SERVICE

Academic Reviewing

I serve as a reviewer for leading venues across computer vision, multimedia, machine learning, and AI systems.

IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)
IEEE Transactions on Multimedia (TMM)
IEEE Transactions on Image Processing (TIP)
IEEE Transactions on Circuits and Systems for Video Technology (TCSVT)
European Conference on Computer Vision (ECCV)
Conference on Neural Information Processing Systems (NeurIPS)
Conference on Computer Vision and Pattern Recognition (CVPR)
International Conference on Machine Learning (ICML)

03 · TRAINING

Education

2025.10 – Present
Ph.D. in Computer Science and Technology

The University of Sydney · Advisor: Chang Xu

Research: Embodied AI, VLA/WAM architecture, reasoning and acceleration, image enhancement, and machine learning.

2021.09 – 2025.06
B.Eng. in Computer Science and Technology

Northwestern Polytechnical University · Jichang Kang Intelligent Systems Elite Class

04 · EXPERIENCE

Internship

2025.10 – 2026.04
Research Intern · Stellar Edge AI, Hangzhou

Built an end-to-end VLA deployment pipeline for humanoid robots and Franka/SO100 arms; designed pruning and quantization methods that reduced end-to-end latency by about 40%.

2024.07 – 2024.08
Graphics Algorithm Intern · Zhongjin Huian (Beijing) Technology Co., Ltd.

Developed low-light enhancement and real-time video analysis modules for bank-counter surveillance, improving violation detection accuracy by more than 15%.

2024.01 – 2024.02
Scheduling Algorithm Architect Assistant · Beijing Hangzhen Technology Co., Ltd.

Beijing City, China

Assisted the intelligent production-line scheduling algorithm architect in introducing AI large models for architecture design and algorithm optimization.

05 · RECOGNITION

Honors and Awards