Henggang Cui

Picture of Henggang Cui Ph.D.
Uber Advanced Technologies Group
Parallel Data Lab
Department of Electrical and Computer Engineering
Carnegie Mellon University


      Email:
cuihenggang (at) gmail (dot) com

About me

I'm a Ph.D. graduated from Carnegie Mellon University, advised by Greg Ganger. My research area was system support for large-scale machine learning.

I am now a Senior Autonomy Engineer at Uber Advanced Technologies Group, working on autonomous self-driving cars!

[resume]

Thesis

Exploiting Application Characteristics for Efficient System Support for Data-Parallel Machine Learning [dissertation] [slides]

Open-sourced Projects

[GeePS]: GPU-specialized parameter server for GPU machine learning

[MLtuner-GeePS]: GeePS with automatic tuning on learning rate, momentum, batch size, data staleness and more

[IterStore]: High-performacne parameter server for iterative convergent machine learning

Publications

Deep Kinematic Models for Physically Realistic Prediction of Vehicle Trajectories [link]
Henggang Cui, Thi Nguyen, Fang-Chieh Chou, Tsung-Han Lin, Jeff Schneider, David Bradley, Nemanja Djuric
arXiv preprint 1908.00219

Multimodal Trajectory Predictions for Autonomous Driving using Deep Convolutional Networks [link] [pdf]
Henggang Cui, Vladan Radosavljevic, Fang-Chieh Chou, Tsung-Han Lin, Thi Nguyen, Tzu-Kuo Huang, Jeff Schneider, Nemanja Djuric
IEEE International Conference on Robotics and Automation, 2019 (ICRA'19)

Predicting Motion of Vulnerable Road Users using High-Definition Maps and Efficient ConvNets [link] [pdf]
Fang-Chieh Chou, Tsung-Han Lin, Henggang Cui, Vladan Radosavljevic, Thi Nguyen, Tzu-Kuo Huang, Matthew Niedoba, Jeff Schneider, Nemanja Djuric
NeurIPS Workshop on Machine Learning for Intelligent Transportation Systems, 2018 (NeurIPS'18 MLITS workshop)

Short-term Motion Prediction of Traffic Actors for Autonomous Driving using Deep Convolutional Networks [link]
Nemanja Djuric, Vladan Radosavljevic, Henggang Cui, Thi Nguyen, Fang-Chieh Chou, Tsung-Han Lin, Jeff Schneider
arXiv preprint 1808.05819

MLtuner: System Support for Automatic Machine Learning Tuning [fork me at Github]
Henggang Cui, Gregory R. Ganger, and Phillip B. Gibbons
[older PDL TR version] CMU Parallel Data Lab Technical Report 2016 [pdf]
[newer arxiv version] arXiv preprint 1803.07445 [link] [pdf]

Addressing the Straggler Problem for Iterative Convergent Parallel ML [link]
Aaron Harlap, Henggang Cui, Wei Dai, Jinliang Wei, Gregory R. Ganger, Phillip B. Gibbons, Garth A. Gibson, and Eric P. Xing
ACM Symposium on Cloud Computing 2016 (SoCC'16)

GeePS: Scalable Deep Learning on Distributed GPUs with a GPU-Specialized Parameter Server [link] [pdf] [fork me at Github]
Henggang Cui, Hao Zhang, Gregory R. Ganger, Phillip B. Gibbons, and Eric P. Xing
ACM European Conference on Computer Systems, 2016 (EuroSys'16)

Using Data Transformations for Low-latency Time Series Analysis [link] [pdf] [full version]
Henggang Cui, Kimberly Keeton, Indrajit Roy, Krishnamurthy Viswanathan, and Gregory R. Ganger
ACM Symposium on Cloud Computing, 2015 (SoCC'15)

Managed Communication and Consistency for Fast Data-Parallel Iterative Analytics [link]
Jinliang Wei, Wei Dai, Aurick Qiao, Qirong Ho, Henggang Cui, Gregory R. Ganger, Phillip B. Gibbons, Garth A. Gibson, and Eric P. Xing
ACM Symposium on Cloud Computing, 2015 (SoCC'15)
Best Paper Award

Exploiting Iterative-ness for Parallel ML Computations [link] [pdf] [fork me at Github]
Henggang Cui, Alexey Tumanov, Jinliang Wei, Lianghong Xu, Wei Dai, Jesse Haber-Kucharsky, Qirong Ho, Gregory R. Ganger, Phillip B. Gibbons, Garth A. Gibson, and Eric P. Xing
ACM Symposium on Cloud Computing, 2014 (SoCC'14)

Exploiting Bounded Staleness to Speed Up Big Data Analytics [link] [pdf]
Henggang Cui, James Cipar, Qirong Ho, Jin Kyu Kim, Seunghak Lee, Abhimanu Kumar, Jinliang Wei, Wei Dai, Gregory R. Ganger, Phillip B. Gibbons, Garth A. Gibson, and Eric P. Xing
USENIX Annual Technical Conference, 2014 (ATC'14)

More Effective Distributed ML via a Stale Synchronous Parallel Parameter Server [link] [pdf]
Qirong Ho, James Cipar, Henggang Cui, Jin Kyu Kim, Seunghak Lee, Phillip B. Gibbons, Garth A. Gibson, Gregory R. Ganger, and Eric P. Xing
Neural Information Processing Systems, 2013 (NIPS'13)

Optically Cross-Braced Hypercube: A Reconfigurable Physical Layer for Interconnects and Server-Centric Datacenters [link] [pdf]
Henggang Cui, Danielle Rasooly, Moises R. N. Ribeiro, and Leonid Kazovsky
Optical Fiber Communication Conference, 2012 (OFC'12)

Scalable Data Center Multicast using Multi-class Bloom Filter [link] [pdf]
Dan Li, Henggang Cui, Yan Hu, Yong Xia, and Xin Wang
19th IEEE International Conference on Network Protocols, 2011 (ICNP'11)

Talks

Exploiting Application Characteristics for Efficient System Support of Data-parallel Machine Learning [slides]
Henggang Cui
CMU Ph.D. Thesis Defense, Pittsburgh, PA, April 2017

MLtuner: System Support for Automatic Machine Learning Tuning [slides]
Henggang Cui, Gregory R. Ganger, and Phillip B. Gibbons
Parallel Data Lab Retreat, Bedford Springs, PA, October 2016

GeePS: Scalable Deep Learning on Distributed GPUs with A GPU-specialized Parameter Server [slides]
Henggang Cui, Hao Zhang, Gregory R. Ganger, Phillip B. Gibbons, and Eric P. Xing
ACM European Conference on Computer Systems (EuroSys'16), London, UK, April 2016

Using Data Transformations for Low-latency Time Series Analysis [slides]
Henggang Cui, Kimberly Keeton, Indrajit Roy, Krishnamurthy Viswanathan, and Gregory R. Ganger
ACM Symposium on Cloud Computing (SoCC'15), Kohala Coast, HI, August 2015

Exploiting Iterative-ness for Parallel ML Computations [slides]
Henggang Cui, Alexey Tumanov, Jinliang Wei, Lianghong Xu, Wei Dai, Jesse Haber-Kucharsky, Qirong Ho, Gregory R. Ganger, Phillip B. Gibbons, Garth A. Gibson, and Eric P. Xing
ACM Symposium on Cloud Computing (SoCC'14), Seattle, WA, November 2014

Exploiting Bounded Staleness to Speed Up Big Data Analytics [slides]
Henggang Cui, James Cipar, Qirong Ho, Jin Kyu Kim, Seunghak Lee, Abhimanu Kumar, Jinliang Wei, Wei Dai, Gregory R. Ganger, Phillip B. Gibbons, Garth A. Gibson, and Eric P. Xing
USENIX Annual Technical Conference (ATC'14), Philadelphia, PA, June 2014

LazyTable: Distributed Machine Learning with the Stale Synchronous Parallel Model [slides]
Henggang Cui, James Cipar, Qirong Ho, Jin Kyu Kim, Abhimanu Kumar, Seunghak Lee, Greg R. Ganger, Phil B. Gibbons, Garth A. Gibson, and Eric P. Xing
WIP Talk, ACM Symposium on Operating Systems Principles (SOSP'13), Farmington, PA, November 2013

Public services

Reviewer of IEEE International Conference on Robotics and Automation 2019. [link]
Reviewer of IEEE Robotics and Automation Letters 2019.
Reviewer of IEEE Big Data 2019. [link]
Program committee of NeurIPS 2019 "Machine Learning for Autonomous Driving" Workshop. [link]
Reviewer of Cluster Computing 2019.
Program committee of CVPR 2019 "Precognition: Seeing through the Future" Workshop. [link]
Reviewer of IEEE Transactions on Network Science and Engineering 2018.
Program committee of ACM Symposium on Cloud Computing 2018. [link]
Program committee of Computing Conference 2018. [link]
Program committee of ACM Symposium on Cloud Computing 2017. [link]

Teaching

Teaching assistant of Storage Systems (15746/18746), Spring 2015. [link]
Teaching assistant of Storage Systems (15746/18746), Fall 2016. [link]

Misc

Henggang Cui at [LinkedIn], [Google Scholar], [Github], [iNaturalist], [eBird], [Flickr].