Since 2021, I have been a PhD student at HKUST co-advised by Lionel Ni (President of HKUST-GZ) and Harry Shum (Former Executive Vice President of Microsoft). My research focuses on image and video generation, visual editing, and talking head synthesis.
From May 2026, I will join Meta as a research intern, working on video generation with Lu Yuan. Previously, from May 2025 to April 2026, I interned at IDEA and StepFun, collaborating with Lei Zhang and Gang Yu, and co-authored nine papers on image/video generation and editing. Concurrently, I led R&D teams working on video generation and agents at Xiaobing.ai.
From April 2023 to April 2025, I co-founded Morph Studio, a video generation startup with over 1.5 million users. Earlier, starting in 2022, I worked as a research intern at Xiaobing.ai, collaborating with Baoyuan Wang and Duomin Wang.
From August 2019 to May 2021, I worked with Carlo H. Séquin at UC Berkeley on the graphics project JIPCAD.
I received my B.S. from the Department of Computer Science and Technology (Honors Science Program) and completed the Honors Youth Program (少年班) at Xi'an Jiaotong University in 2021.
2026 IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2026,
We show that unlocking MLLM reasoning via a thinking–editing–reflection loop significantly improves image editing accuracy by enhancing instruction understanding and automatic error correction, achieving consistent gains over state-of-the-art diffusion-based editors such as Step1X-Edit and Qwen-Image-Edit.
The Fourteenth International Conference on Learning Representations, ICLR 2026,
Completed from scratch in 30 days
We introduce ColorCtrl, a training-free method for text-guided color editing in images and videos. It enables precise, word-level control of color attributes while preserving geometry and material consistency. Experiments on SD3, FLUX.1-dev, and CogVideoX show that ColorCtrl outperforms existing training-free and commercial models, including GPT-4o and FLUX.1 Kontext Max, and generalizes well to instruction-based editing frameworks.
The Fourteenth International Conference on Learning Representations, ICLR 2026,
Completed from scratch in 30 days
LazyDrag, a drag-based image editing method for Multi-Modal Diffusion Transformers, eliminates implicit point matching, enabling precise geometric control and text guidance without test-time optimization.
The Fourteenth International Conference on Learning Representations, ICLR 2026,
We introduce SpeakerVid-5M, the first large-scale dataset designed specifically for the audio-visual dyadic interactive virtual human task.
ACM SIGGRAPH Conference and Exhibition on Computer Graphics and Interactive Techniques in Asia, ACM SIGGRAPH Asia 2025,
Oral Presentation (Completed from scratch in 21 days)
ConsistEdit is a training-free attention control method for MM-DiT that enables precise, structure-aware image and video editing. It supports multi-round edits with strong consistency and achieves state-of-the-art performance without manual design or test-time tuning.
ACM SIGGRAPH Conference and Exhibition on Computer Graphics and Interactive Techniques in Asia, ACM SIGGRAPH Asia 2025,
Oral Presentation
This work addresses animation retargeting across characters with differing skeletal topologies. Motion2Motion is a novel, training-free framework that requires only a few target motions and sparse bone correspondences. It overcomes topological inconsistencies without large datasets. Extensive evaluations show strong performance in both similar and cross-species settings, with practical applications in user-facing tools. Code and data will be released.
arXiv, 2509.06155,
We introduce UniVerse-1, a unified, Veo-3-like model capable of simultaneously generating coordinated audio and video. To enhance training efficiency, we bypass training from scratch and instead employ a stitching of expertise technique.
2023 IEEE International Conference on Computer Vision, ICCV 2023,
We introduce a simple and novel framework for one-shot audio-driven talking head generation. Unlike prior works that require additional driving sources for controlled synthesis in a deterministic manner, we instead probabilistically sample all the holistic lip-irrelevant facial motions (i.e. pose, expression, blink, gaze, etc.) to semantically match the input audio while still maintaining both the photo-realism of audio-lip synchronization and the overall naturalness.
2023 IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2023,
We present a novel one-shot talking head synthesis method that achieves disentangled and fine-grained control over lip motion, eye gaze&blink, head pose, and emotional expression. We represent different motions via disentangled latent representations and leverage an image generator to synthesize talking heads from them.