Chen (Jeremy) Wang

I am an incoming PhD student in Computer Science, Stanford.

I used to work in Machine Vision and Intelligence Group (MVIG), advised by Prof. Cewu Lu . In Summer 2018, I worked at Stanford Vision and Learning Lab, advised by Prof. Silvio Savarese, Prof. Fei-Fei Li and Prof. Yuke Zhu. During the last semester of my undergraduate study, I was fortunate to work as an intern researcher at CSAIL MIT, advised by Prof. Ted Adelson and Prof. Ruth Rosenholtz.

Email  /  CV  /  Github  /  Google Scholar

Education

Stanford University, CA, USA
Incoming Ph.D. in Computer Science

Shanghai Jiao Tong University, SH, China
B.S. in Computer Science, Sep. 2015 to Jun. 2020
(gapped in 2016 due to health reasons)
Advisor: Prof. Cewu Lu

Internship

Massachusetts Institute of Technology, MA, USA
Research Intern . Sep. 2019 to Mar. 2020
Advisors: Prof. Ted Adelson and Prof. Ruth Rosenholtz

Stanford University, CA, USA
Research Intern . Jun. 2018 to Aug. 2019
Advisors: Prof. Silvio Savarese, Prof. Fei-Fei Li and Prof. Yuke Zhu

Publications
SwingBot: Learning Physical Features from In-hand Tactile Exploration for Dynamic Swing-up Manipulation
Chen Wang*, Shaoxiong Wang*, Branden Romero, Filipe Veiga, Edward Adelson
International Conference on Intelligent Robots and Systems
IROS 2020 Best Paper Award

[paper] [website] [presentation]

6-PACK: Category-level 6D Pose Tracker with Anchor-Based Keypoints
Chen Wang, Roberto Martin-Martin, Danfei Xu, Jun Lv, Cewu Lu , Li Fei-Fei, Silvio Savarese, Yuke Zhu
International Conference on Robotics and Automation, ICRA 2020

[paper] [website] [video] [code]

DenseFusion: 6D Object Pose Estimation by Iterative Dense Fusion
Chen Wang, Danfei Xu, Yuke Zhu, Roberto Martin-Martin, Cewu Lu , Li Fei-Fei, Silvio Savarese
Computer Vision and Pattern Recognition, CVPR 2019

[paper] [website] [video] [code]

Transferable Force-Torque Dynamics Model for Peg-in-hole Task
Junfeng Ding*, Chen Wang*, Cewu Lu
International Conference on Intelligent Robots and Systems, IROS 2019

[paper] [video]

TendencyRL: Multi-stage discriminative hints for efficient goal-oriented reverse curriculum learning
Chen Wang*, Junfeng Ding*, Xiangyu Chen, Zelin Ye, Jialu Wang, Ziruo Cai, Shixiang Gu, Cewu Lu
International Conference on Intelligent Robots and Systems, IROS 2019
[paper] [video]
Deep Reinforcement Learning Symposium, NIPS Symposium 2017
[poster]

Open-sourced Projects

DenseFusion (~650 stars, ~200 forks)
Open-Source to facilitate further research, Jan 2019
• Full implementation code of the paper DenseFusion (CVPR2019)

  • A dense fusion network that combines pixel-wise color and depth information to an integral color-depth feature embedding for 6-DoF object pose estimation.
  • An iterative refinement procedure within the neural network architecture that enhances model performance while keeping the inference speed real-time.

6-PACK (~180 stars)
Open-Source to facilitate further research, Oct 2019
• Full implementation code of the paper 6-PACK (ICRA 2020)

  • An unsupervised 3D keypoints generation network used for category-level 6D object pose tracking. This method can track objects without knowing their mesh models.

SLAM-based 6D pose tracking baseline
Open-Source to facilitate further research, May 2019
•An implementation of a basic model-free rigid object 6D pose tracking baseline with ICP(w/wo color registration) and truncated signed distance function(TSDF) fusion.

MuJoCo-Unity UR5 robot arm simulation
Open-Source to facilitate further research, May 2018
•A simulation environment combined with the Unity rendering engine and MuJoCo physics engine, which can be used as training tools for Reinforcement Learning tasks.


Service

Reviewer:
•IEEE Robotics and Automation Letters (RA-L)
•International Conference on Robotics and Automation (ICRA)
•International Conference on Intelligent Robots and Systems (IROS)


 

Template source