Haoxin Chen
Video Generation · Algorithm Researcher & Engineer
Building the future of AI-generated video. Contributed to
VideoCrafter, HunyuanVideo, and
Seedance. Currently exploring real-time interactive
multimodal generation at vivix.ai.
3,100+ Google Scholar citations · Published at CVPR, ECCV, ICLR, SIGGRAPH Asia.
Currently looking for like-minded partners in real-time interaction and
multimodal generation to collaborate on AI-native algorithm R&D.
Get in touch if you're interested.
I am a video generation algorithm researcher and engineer specializing in diffusion-based generative models. I obtained my Bachelor's and Master's degrees from South China University of Technology (SCUT) (2015 – 2022). I have worked at Tencent (2021–2024) and ByteDance (2024–2026), participating in VideoCrafter, HunyuanVideo, Seedance, and AI-powered short drama platforms (Douyin & Honguo). Currently, I am at vivix.ai, where I work on streaming-native, multimodal real-time interactive generation through Vivix-A1 and Vivix-W1.
Algorithm Researcher & Engineer
Streaming-native, multimodal real-time interactive generation R&D, including Vivix-A1 and Vivix-W1.
Algorithm Researcher & Engineer
Participated in Seedance video generation model R&D. Worked on algorithm systems for Douyin and Honguo AI short drama platforms.
Algorithm Researcher & Engineer
Participated in the VideoCrafter series and HunyuanVideo — open-source video generation foundation models.
B.Eng & M.Eng
Bachelor's and Master's degrees. Research on computer vision, video understanding, and video object segmentation.
Vivix-A1
Real-time, full-duplex AI characters powered by low-latency streaming audio-video generation for continuous interaction in open worlds.
Vivix-W1
A streaming-native world model with text, image, audio, and video references that can keep shaping an evolving audio-visual world during generation.
HunyuanVideo
Video generation model developed by Tencent.
Seedance
Video generation model developed by ByteDance.
Video Diffusion — Data & Training Recipe
Focused on overcoming data limitations and establishing open-source foundations for high-quality video diffusion models.
-
H Chen, Y Zhang, X Cun, M Xia, X Wang, C Weng, Y ShanCVPR2024705 cited
-
H Chen, M Xia, Y He, Y Zhang, X Cun, S Yang, J Xing, Y Liu, Q Chen, ...arXiv2023638 cited
Image-to-Video & Animation
Exploring how to breathe life into static images through video diffusion priors and controllable generation.
-
J Xing, M Xia, Y Zhang, H Chen, X Wang, TT Wong, Y ShanECCV2023541 cited
-
J Xing, M Xia, Y Liu, Y Zhang, Y Zhang, Y He, H Liu, H Chen, X Cun, ...TVCG2024168 cited
-
Y He, M Xia, H Chen, X Cun, Y Gong, J Xing, Y Zhang, X Wang, C Weng, ...arXiv2023114 cited
-
Y He, S Yang, H Chen, X Cun, M Xia, Y Zhang, X Wang, R He, Q Chen, ...ICLR2023135 cited
Evaluation & Benchmark
Building rigorous evaluation frameworks for video generation models.
-
Y Liu, X Cun, X Liu, X Wang, Y Zhang, H Chen, Y Liu, T Zeng, R Chan, ...CVPR2024385 cited
Video Object Segmentation — Earlier Work
Master's research on few-shot video understanding with attention-based architectures.
-
H Chen, H Wu, N Zhao, S Ren, S HeCVPR202130 cited
Open to discussions about collaboration, internships, and job opportunities. Feel free to reach out via email.