- 🤓 Working on something new; reach out to learn more
👨💻 Ex-ML Engineering Lead @ToyotaResearch
🤖 All about self-supervised pre-training of robots
🧠 AVs/Drones, ML/DL, Computer Vision, Visual SLAM
- 💻 LinkedIn » Twitter » Github » GScholar » CV » Email
Present: I am currently working something new; stay tuned!
2019 - 2021: I led the ML Engineering team at Toyota Research Institute (TRI) that was responsible for the design and development of all ML models deployed into Toyota’s next-generation L4 autonomous vehicle fleet. The work developed here was most recently presented in Ryan Eustice’s SPIE Future Sensing Technologies 2020 keynote and at the Pytorch Ecosystem Day 2021 poster session.
- 2017 - 2019: I spent my initial tenure at Toyota Research Institute (TRI) as a research scientist in the Machine Learning team, spearheading their large-scale self-supervised learning efforts for autonomous driving. For a detailed set of publications and patents stemming from this work, please refer below to the academic publications section.
2011 - 2017: My PhD at MIT CSAIL focused on self-supervised perception and learning in SLAM-aware mobile robots. I was advised by John Leonard with my thesis committee comprising of Nick Roy, Leslie Kaelbling and Antonio Torralba. My thesis made key contributions in monocular SLAM supported 3D object recognition, self-supervised visual ego-motion learning, and self-supervised visual place recognition in spatially-cognizant mobile robots. I also spent a summer interning at Mitsubishi Electric Research Labs (MERL) developing high-performance 3D stereo reconstruction techniques with Srikumar Ramalingam.
2009 - 2011: After undergrad, I spent two years at PhaseSpace Motion Capture where I worked as a computer vision engineer developing cutting-edge 3D computer vision and motion capture technologies.
- 2005 - 2009: BS at the University of Michigan - Ann Arbor. I was a big gear-head at the time, working on Formula SAE and co-founded UM::Autonomy.
|2017 - 2021||2011 - 2017||2014||2009 - 2011||2007||2005 - 2008|
See Google Scholar for the most up-to-date publications and patents list.
Neural Ray Surfaces for Self-Supervised Learning of Depth and Ego-motion
3DV 2020 Igor Vasiljevic, Vitor Guizilini, Rares Ambrus, Sudeep Pillai, Wolfram Burgard, Greg Shakhnarovich, Adrien Gaidon
PillarFlow: End-to-end Birds-eye-view Flow Estimation for Autonomous Driving
IROS 2020 Kuan-Hui Lee, Matthew Kliemann, Adrien Gaidon, Jie Li, Chao Fang, Sudeep Pillai, Wolfram Burgard
Self-Supervised 3D Keypoint Learning for Ego-motion Estimation
CoRL 2020 Jiexiong Tang, Rares Ambrus, Vitor Guizilini, Sudeep Pillai, Hanme Kim, Patric Jensfelt, Adrien Gaidon
PackNet-SfM: 3D Packing for Self-Supervised Monocular Depth Estimation
CVPR 2020 Vitor Guizilini, Rares Ambrus, Sudeep Pillai, Allan Raventos, Adrien Gaidon
Robust Semi-Supervised Monocular Depth Estimation with Reprojected Distances
CoRL 2019 Vitor Guizilini, Jie Li, Rares Ambrus, Sudeep Pillai, Adrien Gaidon
Neural Outlier Rejection for Self-Supervised Keypoint Learning
CoRL 2019 Jiexiong Tang, Hanme Kim, Vitor Guizilini, Sudeep Pillai, Rares Ambrus
Two Stream Networks for Self-Supervised Ego-Motion Estimation
CoRL 2019 Rares Ambrus, Vitor Guizilini, Jie Li, Sudeep Pillai, Adrien Gaidon
SuperDepth: Self-Supervised, Super-Resolved Monocular Depth Estimation
ICRA 2019 Sudeep Pillai, Rares Ambrus, Adrien Gaidon
SLAM-Aware, Self-Supervised Perception in Mobile Robots
PhD Thesis 2017 Sudeep Pillai
Exploring big volume sensor data with Vroom
VLDB Demo 2017 Oscar Moll, Aaron Zalewski, Sudeep Pillai, Sam Madden, Michael Stonebraker, Vijay Gadepally
Self-Supervised Visual Place Recognition Learning in Mobile Robots
IROS Workshop 2017 Sudeep Pillai, John Leonard
Towards visual ego-motion learning in robots
IROS 2017 Sudeep Pillai, John Leonard
SLAMinDB: Centralized graph databases for mobile robotics
ICRA 2017 Dehann Fourie, Sam Classens, Sudeep Pillai, Roxana Mata, John Leonard
High-Performance and Tunable Stereo Reconstruction
ICRA 2016 Sudeep Pillai, Srikumar Ramalingam, John Leonard
Monocular SLAM Supported Object Recognition
RSS 2015 Sudeep Pillai, John Leonard
Line-sweep: Cross-ratio for wide-baseline matching and 3D reconstruction
CVPR 2015 Srikumar Ramalingam, Michel Antunes, Dan Snow, Gim Hee Lee, Sudeep Pillai
Bitcoin Transaction Graph Analysis
arXiv 2013 Michael Fleder, Michael Kester, Sudeep Pillai
- TWIMLcon 2021: MLOps for High-Stakes Environments
- ODSC West 2019: World-scale Deep Learning for Automated Driving
- Auto AI 2019: World-scale Deep Learning for Automated Driving
- NVIDIA GTC Silicon Valley 2019: Beyond Supervised Driving (jointly with Adrien Gaidon)
- RE·WORK Deep Learning Summit 2018: Self-supervision in Mobile Robots in the Deep Learning Era