Published several papers, including 'Shape My Moves: Text-Driven Shape-Aware Synthesis of Human Motions' (CVPR 2025), 'DepthScape: Authoring 2.5D Designs with Depth Estimation' (CHI 2025), 'GUI-Bee: Align GUI Action Grounding to Novel Environments via Autonomous Exploration' (arXiv 2025), 'Copiloting Creative 3D Scene Modeling and Visualization with Generative Agents' (NeurIPS 2024), and 'GPU-accelerated Lossless Image Compression with Massive Parallelization' (ISM 2023).
Research Experience
Currently a research scientist at Adobe Research. Previously worked for Tencent, DJI, and Hiscene.
Education
Ph.D. in Computer Science from the University of Maryland, advised by Prof. Ming Lin.
Background
Research interests include MLLM-based copilot and image compression, such as Project Scenic presented in Adobe MAX 2024. His research experiences cover LLM, computer vision, 3D perception, multi-modality learning, robust learning, etc., in real-world applications like MLLM-based copilot, AR/VR/Metaverse, autonomous driving, virtual try-on, etc.