Enjoyed giving a keynote speech at CVPR Workshop on Autonomous Driving, sharing our recent results learning complex multi-agent interaction behaviours for driving in busy urban environments. Recording is here: youtu.be/Va-F4qtTQ6g

Excellent work from Lu Mi's internship at Waymo!

Our team got the first place in the Waymo Motion Prediction Challenge! I know this is not easy for a college team, credit goes to my amazing students! Our winning method is called DenseTNT, more information will be released soon.

Excited to be involved in 3 papers at CVPR'21: - Purely perspective 3D detection in point cloud: tinyurl.com/y998vmpc - 2-stage perspective 3D detection in point cloud: tinyurl.com/5dp6mv2w - HDMap generation: tinyurl.com/2ybtr5n5

Our work on Learning Animatable 3D Human Shape and Skeleton in a Single Neural Network.

Benchmarking Representation Learning for Natural World Image Collections. See how SimCLR, SwAV, and MoCo compare to supervised baselines when trained on the new iNat2021 dataset. Come chat with me @ #CVPR2021 #CVPR June 24, 2021 11AM-1:30PM EDT (sess 10)!

Is monocular 3D perception needed? Is the field making meaningful progress? What do you think? Tell us below and join our Mono3D #CVPR2021 workshop on Friday 25 8am-12pm PT sites.google.com/view/mono3d-wo…! We have great speakers, 2 panels, and we'll announce the DDAD challenge winners!

Excited to be in Phila and have to opportunity to present alongside so many robotics visionaries at the Scaling Robot Learning workshop (sites.google.com/view/icra22-srl) at ICRA'22 tomorrow! #ICRA2022 #selfDrivingCar #cruise

Excited to be at the first "real" conference in 2.5 years! Hit me up if you are in New Orleans the coming week! #CVPR22

Interested in our ongoing effort in deploying driverless vehicle in SF and beyond and happen to be in New Orleans by Chance? Come join my talk at cvpr2022.wad.vision on Monday: ⏰: 415pm 🛖: Great Hall C #cruise #CVPR2022 #SelfDrivingCars

Best party in town at barcaria tonight! #cruise #CVPR22

Had a full in-person audience + hundreds more on zoom listening to my talk about #Cruise at the WAD workshop at #CVPR22. Recording to follow soon! #SelfDrivingCars

An incredible line-up of confirmed speakers: Abhinav Valada Andrew Davison @HLategahn Philipp Krähenbühl Yuning Chai 💡 More information: sites.google.com/view/mlad-eccv…

I was privileged to represent #cruise at the self-driving workshop at #cvpr2022, where I highlighted some of our long-term investments (uncertainty, radar, RL, etc.) and discussed the challenges to deploy the AV. Check out the newly released recording: youtube.com/watch?v=EJ0hgO…

Wanna switch from academia to industry but unsure how? Here's a guide! rowanmcallister.github.io/post/industry/ Thanks to Greg Kahn, Kate Rakelly, Boris Ivanovic, Rebekah Baratho, Ashwin Balakrishna, Jessica Yin, Nick Rhinehart, Jessica Cataneo, Rareș Ambruș

📜🚨📜🚨 NN loss landscapes are full of permutation symmetries, ie. swap any 2 units in a hidden layer. What does this mean for SGD? Is this practically useful? For the past 5 yrs these Qs have fascinated me. Today, I am ready to announce "Git Re-Basin"! arxiv.org/abs/2209.04836

We are open for business for 2023 interns! Our team works on projects across the self-driving stack, from perception, prediction, planning to simulation and testing. If you are interested, please reach out to me, or apply here: getcruise.com/careers/jobs/4…

At CVPR this week. Excited to present our research work at Cruise at the E2E AD workshop (opendrivelab.com/e2ead/cvpr23.h…) on Sunday and the SAIAD workshop (sites.google.com/view/saiad2023…) on Monday. See you there!

(1/N ) Can a large multimodal model not only understand bboxes, but also understand arbitrary visual prompts (scribble, arrow, etc) without explicit region embedding? Yes! Our latest work ViP-LLaVA arxiv.org/abs/2312.00784 shows an extremely simple but effective approach.