Yixiao Ge
Head Of Embodied AI And Data
Beijing, China
Summary
🤩
Rockstar🎓
Top SchoolYixiao Ge is a research-driven AI leader with 11 years' experience building vision and multimodal foundation models, currently founding the Multimodal Intelligence Department at XPENG Robotics to advance embodied multimodal large models and world models. Previously a Principal Researcher at Tencent ARC Lab, he led research on vision and multimodal systems, progressing through senior research roles that combined deep learning theory with product-relevant engineering. He holds a PhD in Computer Vision and Deep Learning from The Chinese University of Hong Kong and a bachelor’s in Automation from Huazhong University of Science and Technology. Yixiao blends rigorous academic training with hands-on lab-to-product experience across industry research labs in Beijing and Hong Kong. Colleagues know him for turning complex perception problems into scalable model architectures and system prototypes that drive robotics capabilities. He maintains an active professional presence through his homepage and GitHub, reflecting ongoing work at the intersection of embodied AI and large multimodal models.
11 years of coding experience
6 years of employment as a software developer
Bachelor's degree Automation, Bachelor's degree Automation at Huazhong University of Science and TechnologyThe Chinese University of Hong Kong (CUHK)
Github Skills (45)
cross-domain10
re-identification10
image-translation10
image-retrieval10
localization10
unsupervised-learning10
adversarial-learning10
contrastive-learning10
toolbox9
pytorch9
semi-supervised-learning9
paper9
few-shot-learning9
adaptation9
transfer-learning9
Programming languages (4)
Github contributions (5)
yxgeee/MMT
Jan 2020 - Jul 2020
[ICLR-2020] Mutual Mean-Teaching: Pseudo Label Refinery for Unsupervised Domain Adaptation on Person Re-identification.
Contributions:19 commits, 1 PR, 18 pushes in 5 months
re-identificationunsupervised-domain-adaptationperson-re-identificationunsupervised-learningpseudo-labels
open-mmlab/OpenUnReID
Jul 2020 - Jun 2021
PyTorch open-source toolbox for unsupervised or domain adaptive object re-ID.
Contributions:7 reviews, 70 commits, 5 PRs in 11 months
pytorchunsupervised-learningunsupervised-domain-adaptationre-identificationimage-retrieval
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.