Computer Vision / Deep Learning / AI @ Samsara | Ex-Amazon
Email: kevin.lai.726 "at" gmail dot com
I am currently an Applied Science & Engineering Manager at Samsara leading a team to build and deploy state-of-the-art Computer Vision and Large Vision Language Models (LVLM / LLM) on cloud and on edge compute for improving the safety, sustainability, and efficiency of physical operations. Road-facing features include lane departure, following distance, collision warning, and more. Driver-facing features include drowsiness, mobile usage, seat belt usage, and more (U.S. Patents 12165393, 12260616, 12266123, 12272138, 12327417).
Previously I led a team of applied scientists and machine learning engineers on Amazon's physical stores tech team that delivered the Just Walk Out cashier-less checkout technology. By building and deploying computer vision and deep learning to the cloud and to edge compute, we created a store where customers can simply take what they want and go! No lines, no checkout. Today these systems run in over 100 Just Walk Out stores operated by Amazon and third parties who license Amazon's technology.
I completed my Ph.D. in Computer Science at the University of Washington in Seattle with Professor Dieter Fox in Dec 2013. Before that, I obtained my B.Sc. in Computer Science at the University of British Columbia in Vancouver, BC.
Drowsy Driving Detection
Sung Chun Lee, Nathan Hurst, Yan Wang, Olamide Akintewe, Justin Levine, Kenshiro Nakagawa, Cole Jurden, Rachel Demerly, Aravindh Ramesh, Kevin Lai, Jovanna Bubar, Shirish Nair, Maisie Wang
U.S. Patent 12 327 417, 2025.
Forward Collision Warning
Rohit Annigeri, Sharan Srinivasan, Kevin Lai, Jose Cazarin, Brian Westphal, Shiva Bala, Ivan Stoev, Douglas Boyle, Cole Jurden, Margaret Irene Finch, Rachel Demerly, Maya Krupa, Shirish Nair, Nathan Hurst, Yan Wang, Shaurye Aggarwal, Akshay Raj Dhamija
U.S. Patent 12 272 138, 2025.
Monitoring the Safe Distance Between Vehicles While Driving
Suryakant Kaushik, Cole Jurden, Marc Clifford, Robert Koenig, Abner Ayala, Kevin Lai, Jose Cazarin, Margaret Irene Finch, Rachel Demerly, Nathan Hurst, Yan Wang, Akshay Raj Dhamija
U.S. Patent 12 266 123, 2025.
Multi-task Machine Learning Model for Event Detection
Narendran Rajan, Yan Wang, Phil Ammirato, Kevin Lai, Evan Welbourne, Nathan Hurst
U.S. Patent 12 260 616, 2025.
Lane Departure Monitoring
Akshay Raj Dhamija, Abner Ayala, Rohit Annigeri, Cole Jurden, Douglas Boyle, Jason Liu, Kevin Lai, Jose Cazarin, Pang Wu, Nathan Hurst, Brian Westphal, Lucas Doyle, Saurabh Tripathi, Shirish Nair
U.S. Patent 12 165 393, 2024
Disambiguating Between Multiple Users
Sudarshan Narasimha Raghavan, Emilio Ian Maldonado, David Allen Smith, Min Xu, Nishitkumar Ashokkumar Desai, Daniel Bibireata, Kevin Kar Wai Lai, Pahal Kamlesh Dalal
U.S. Patent 10 552 750, 2020.
Unsupervised Feature Learning for 3D Scene Labeling. Kevin Lai, Liefeng Bo, and Dieter Fox. IEEE International Conference on Robotics and Automation (ICRA), May 2014. [PDF] [bibtex] [video]
Finalist for Best Vision Paper Award
RGB-D Object Recognition: Features, Algorithms, and a Large Scale Benchmark. Kevin Lai, Liefeng Bo, Xiaofeng Ren, and Dieter Fox. Consumer Depth Cameras for Computer Vision: Research Topics and Applications, 2013. [LINK] [bibtex]
Detection-based Object Labeling in 3D Scenes. Kevin Lai, Liefeng Bo, Xiaofeng Ren, and Dieter Fox. IEEE International Conference on Robotics and Automation (ICRA), May 2012. [PDF] [bibtex] [video] [code]
A Scalable Tree-based Approach for Joint Object and Pose Recognition. Kevin Lai, Liefeng Bo, Xiaofeng Ren, and Dieter Fox. Twenty-Fifth Conference on Artificial Intelligence (AAAI), August 2011. [PDF] [bibtex] [slides]
Object Recognition with Hierarchical Kernel Descriptors. Liefeng Bo, Kevin Lai, Xiaofeng Ren, and Dieter Fox. IEEE International Conference on Computer Vision and Pattern Recognition (CVPR), June 2011. [PDF] [bibtex]
Sparse Distance Learning for Object Recognition Combining RGB and Depth Information. Kevin Lai, Liefeng Bo, Xiaofeng Ren, and Dieter Fox. IEEE International Conference on Robotics and Automation (ICRA), May 2011. [PDF] [bibtex] [slides]
Best Vision Paper Award
A Large-Scale Hierarchical Multi-View RGB-D Object Dataset. Kevin Lai, Liefeng Bo, Xiaofeng Ren, and Dieter Fox. IEEE International Conference on Robotics and Automation (ICRA), May 2011. [PDF] [bibtex] [poster] [dataset]
Object Recognition in 3D Point Clouds Using Web Data and Domain Adaptation. Kevin Lai and Dieter Fox. International Journal of Robotics Research 29(8), Jul 2010. [PDF] [bibtex]
3D Laser Scan Classification Using Web Data and Domain Adaptation. Kevin Lai and Dieter Fox. Robotics: Science and Systems (RSS), Jul 2009. [PDF] [bibtex] [poster]
Curious George: An Attentive Semantic Robot. David Meger, Per-Erik Forssén, Kevin Lai, Scott Helmer, Sancho McCann, Tristram Southey, Matthew Baumann, James J. Little, and David G. Lowe. Robotics and Autonomous Systems Journal 56(6), Jun 2008. [PDF] [bibtex]
Informed Visual Search: Combining Attention and Object Recognition. Per-Erik Forssén, David Meger, Kevin Lai, Scott Helmer, James J. Little, and David G. Lowe. IEEE International Conference on Robotics and Automation (ICRA) 2008, May 2008. [PDF] [bibtex]
Curious George: An Attentive Semantic Robot. David Meger, Per-Erik Forssén, Kevin Lai, Scott Helmer, Sancho McCann,Tristram Southey, Matthew Baumann, James J. Little, David G. Lowe, and Bruce Dow. IROS 2007 Workshop: From Sensors to Human Spatial Concepts, Nov 2007. [PDF] [bibtex]
Curious George: The UBC Semantic Robot Vision System. Scott Helmer, David Meger, Per-Erik Forssén, Sancho McCann, Tristram Southey, Matthew Baumann, Kevin Lai, Bruce Dow, James J. Little, and David G. Lowe. AAAI-07 Mobile Robot Workshop Technical Report, AAAI-WS-07-15, Oct 2007. [PDF] [bibtex]
Object Recognition and Semantic Scene Labeling for RGB-D Data. Kevin Kar Wai Lai. Ph.D. Dissertation, Dec 2013. [PDF] [bibtex]
April 5, 2014 - The RGB-D Scenes Dataset v.2 is now available here. It contains 14 new scenes reconstructed from RGB-D videos with furniture and tabletop objects, as well as Trimble 3D Warehouse objects used to learn HMP3D features and classifiers in our ICRA'14 paper!
March 1, 2014 - I joined Amazon as a Research Scientist.
Nov 22, 2013 - I successfully defended my PhD thesis: Object Recognition and Semantic Scene Labeling using RGB-D Data!
July 14, 2013 - Code for HMP features now available here. It achieves state-of-the-art results on the RGB-D Object Dataset!
December 13, 2012 - Software and data for detection-based object labeling in Kinect videos now available here.
October 3, 2012 - The RGB-D Object Dataset is now available for download directly from the website! No more sending emails necessary (questions and suggestions are, of course, still welcomed!).
May 12, 2011 - We won the ICRA 2011 Best Vision Paper Award! Sparse Distance Learning for Object Recognition Combining RGB and Depth Information
Feb 22, 2011 - The RGB-D Object Dataset is now available! This is a large dataset of 300 objects recorded using a Kinect style 3D camera.