Jianqiang Ren

Jianqiang Ren (任健强)

I am currently a Senior Algorithm Expert at Alibaba Tongyi Lab. I have been actively involved in projects including Luban Intelligent Design, Aliyun Open Vision Platform, Tongyi (Qwen) App, and Wanxiang Digital Human. Before joining Alibaba, I was a VR/AR researcher at NetEase, both as an intern and a full-time staff. I earned my bachelor's and master's degrees from Zhejiang University. My research mainly focuses on human-centric AIGC, multimodal 3D digital avatar generation, animation and humanoid robot control.
E-mail: rjq235@163.com

Scholar / Github

Research

I'm interested in computer vision, human-centric generative AI, multimodal digital avatar, and embodied intelligence.

Selected Publications

	OMG-Avatar: One-shot Multi-LOD Gaussian Head Avatar Jianqiang Ren, Lin Liu, Steven Hoi CVPR, 2026 arXiv / project One-shot Multi-LOD Gaussian head reconstruction from a single image in 0.2s, supporting diverse hardware capabilities and inference speed requirements.
	Mitigating Error Accumulation in Co-Speech Motion Generation via Global Rotation Diffusion and Multi-Level Constraints Xiangyue Zhang, Jianfang Li, Jianqiang Ren, Jiaxu Zhang AAAI, 2026 arXiv A diffusion-based framework operating in global joint rotation space to alleviate hierarchical error accumulation, improving performance by 46% compared to SOTA.
	EchoMask: Speech-Queried Attention-based Mask Modeling for Holistic Co-Speech Motion Generation Xiangyue Zhang, Jianfang Li, Jiaxu Zhang, Jianqiang Ren, Liefeng Bo, Zhigang Tu ACM MM, 2025 arXiv / project A speech-queried attention-based mask modeling framework that selectively masks rhythm-related and semantically expressive motion frames for co-speech motion generation.
	Textoon: Generating Vivid 2D Cartoon Characters from Text Descriptions Chao He, Jianqiang Ren, Yuan Dong, Jianjing Xiang, Xiejie Shen, Weihao Yuan, Liefeng Bo arXiv preprint arXiv:2501.10020 arXiv / project / code Generating diverse 2D cartoon characters in Live2D format from text descriptions within one minute.
	SemTalk: Holistic Co-Speech Motion Generation with Frame-level Semantic Emphasis Xiangyue Zhang, Jianfang Li, Jiaxu Zhang, Ziqiang Dang, Jianqiang Ren, Liefeng Bo, Zhigang Tu ICCV, 2025 arXiv / project / code Holistic co-speech motion generation with frame-level semantic emphasis, separately learning base motions and sparse motions with adaptive fusion.
	Make-a-Character: High Quality Text-to-3D Character Generation within Minutes Jianqiang Ren, Chao He, Lin Liu, Jiahao Chen, Yutong Wang, Yafei Song, Jianfang Li, Tangli Xue, Siqi Hu, Tao Chen, Kunkun Zheng, Jianjing Xiang, Liefeng Bo arXiv preprint arXiv:2312.15430 arXiv / project / code Creating lifelike 3D avatars from text descriptions within 2 minutes, enabling easy integration with CG pipelines for dynamic expressiveness.
	A Hierarchical Representation Network for Accurate and Detailed Face Reconstruction from In-the-Wild Images Biwen Lei, Jianqiang Ren, Mengyang Feng, Miaomiao Cui, Xuansong Xie CVPR, 2023 arXiv / project / code Hierarchical representation network for accurate and detailed face reconstruction with geometry disentanglement and 3D priors of facial details.
	Structure-Aware Flow Generation for Human Body Reshaping Jianqiang Ren, Yuan Yao, Biwen Lei, Miaomiao Cui, Xuansong Xie CVPR, 2022 arXiv / code End-to-end flow generation for body reshaping under guidance of structural priors, achieving controllable performance under arbitrary poses and garments.
	Attention-Aware Multi-Stroke Style Transfer Yuan Yao, Jianqiang Ren, Xuansong Xie, Weidong Liu, Yong-Jin Liu, Jun Wang CVPR, 2019 arXiv / code Integrating self-attention mechanism into style transfer to coordinate spatial attention and render diverse brush strokes harmoniously.

Thanks to Barron's for the template code.