Dinghao YANG   杨丁豪

Senior System Engineer, NVIDIA


Hangzhou, China

Email: dinghaoy@nvidia.com;
             dinghowyang@gmail.com;
Google Scholar: Google Scholar Link
Github: https://github.com/Dinghow
Personal Blog: https://dinghow.site/dinghow-blog

Biography

I am a senior system engineer at NVIDIA, focusing on the training algorithm exploration and framework development for GenAI models, e.g., LLM, Diffusion model. I'm lucky to work with Jiaxin Cao, Junjie Bai and many other talented geeks. Before that, I worked at Lepton AI, and the Machine Learning Platform for AI, Alibaba Cloud, majored in the development of LLM inference framework (I used to be a core developer for the inference engine of Qwen series). I received the M.S degree from the School of Electronic and Computer Engineering, Peking University, in 2023, and the B.E. degree from the School of Software Engineerning, Tongji University, in 2020. I used to be a research intern at Data and Computing Platform, SenseTime Research, during 2020-2022, supervised by Bin Wang and Weijia Li.

My research interests include AI infra (training/inference optimization), point cloud analysis, and image segmentation/matting.

Experience

Publications

Preprint Papers

Cosmos-reason1: From physical common sense to embodied reasoning
NVIDIA (I am a core contributor for the SFT & RL training framework)
[PDF] [Code]
Unified Interactive Image Matting
Dinghao Yang, Bin Wang, Weijia Li, Yiqi Lin, Conghui He
Under Review
[PDF] [Code]

Published Papers

PointCHD: A Point Cloud Benchmark for Congenital Heart Disease Classification and Segmentation
Dinghao Yang, Wei Gao
IEEE Journal Of Biomedical & Health Informatics (JBHI)
[PDF] [Code]
Exploring the User Guidance for More Accurate Building Segmentation from High-Resolution Remote Sensing Images
Dinghao Yang, Bin Wang, Conghui He, Weijia Li
International Journal of Applied Earth Observation and Geoinformation (JAG)
[PDF] [Code]
Exploiting Manifold Feature Representation for Efficient Classification of 3D Point Clouds
Dinghao Yang, Wei Gao, Ge Li, Hui Yuan, Junhui Hou, Sam Kwong
ACM Transactions on Multimedia Computing, Communications and Applications (TOMM)
[PDF] [Code]

Projects

BladeLLM
BladeLLM is a high-performance LLM inference engine of Alibaba Cloud, used for the deployment of Qwen series. I am responsible for the modeling of LLM models, performance optimization, interface development and triton kernel development.
[Introduction]
Intelligent Annotation for SenseBee
I am principally responsible for the interactive data annotation algorithms for SenseBee, the data platform for SenseTime AI research. We propose and deploy deep learning-based algorithms to accelerate manual data annotations (i.e. image segmentation, image matting, 3D object detection).
[Code]
OpenPointCloud
I participate in OpenPointCloud, an open-source algorithm library of deep learning-based point cloud compression & processing, mainly contrbute in the manifold learning-based point cloud representation learning method.
[Code]

Awards

Academic Activities