• At UT-Austin, I look forward to working with strong and motivated students who are passionate about AI + Robotics. If you are interested in joining my lab, please submit your PhD applications to UTCS and mention my name in your statement. Please email me at firstname.lastname@example.org if you have any question.
• At NVIDIA Research, we have year-round internship opportunities in Machine Learning and Robotics. Please contact me at email@example.com for more information.
[new] I will be attending CoRL 2019 and IROS 2019 and giving two invited talks at AnSWeR19 and LRPC workshops.
[new] We released our papers on probabilistic planning under uncertainty and causal reasoning for goal-directed tasks.
[new] We released our technical report on the SURREAL System for distributed reinfrocement learning.
One paper accepted in CoRL 2019 and one paper accepted in NeurIPS 2019.
Two papers accepted in IROS 2019 and one paper accepted in ICCV 2019.
We won the Best Conference Paper Award at ICRA 2019 in Montreal.
Talks and Tutorials
Closing the Perception-Action Loop: Towards General-Purpose Robot Autonomy
Stanford University Ph.D. Dissertation (August 2019)
Learning Keypoint Representations for Robot Manipulation
IROS'19 Workshop on Learning Representations for Planning and Control (November 2019)
6-PACK: Category-level 6D Pose Tracker with Anchor-Based Keypoints
KETO: Learning Keypoint Representations for Tool Manipulation
Causal Induction from Visual Observations for Goal Directed Tasks
Dual Sequential Monte Carlo: Tunneling Filtering and Planning in Continuous POMDPs
Dynamics Learning with Cascaded Variational Inference for Multi-Step Manipulation
CoRL 2019 (Oral Presentation)
Regression Planning Networks
Scaling Robot Supervision to Hundreds of Hours with RoboTurk: Robotic Manipulation Dataset
through Human Reasoning and Dexterity
IROS 2019 (Best Cognitive Robotics Paper Finalist)
DenseFusion: 6D Object Pose Estimation by Iterative Dense Fusion
Neural Task Graphs: Generalizing to Unseen Tasks from a Single Video Demonstration
CVPR 2019 (Oral Presentation)
Making Sense of Vision and Touch: Self-Supervised Learning of Multimodal Representations for
ICRA 2019 (Best Conference Paper)
SURREAL: Open-Source Reinforcement Learning Framework and Robot Manipulation Benchmark
RoboTurk: A Crowdsourcing Platform for Robotic Skill Learning through Imitation
Reinforcement and Imitation Learning for Diverse Visuomotor Skills
Learning Task-Oriented Grasping for Tool Manipulation with Simulated Self-Supervision
Neural Task Programming: Learning to Generalize Across Hierarchical Tasks
Visual Semantic Planning using Deep Successor Representations
Scene Graph Generation by Iterative Message Passing
Target-driven Visual Navigation in Indoor Scenes using Deep Reinforcement Learning
Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotations
Reasoning About Object Affordances in a Knowledge Base Representation