ABSTRACT:

The autonomous vehicle (AV) stack implements a mapping from sensor data to vehicle controls. In this talk I discuss two aspects of this mapping: designing an appropriate objective function and crafting a multimodal object detector. I also discuss nuScenes which is our contribution to the open AV development environment.

BIO:

Oscar Beijom is the Director of Machine Learning at Motional, where he works on products, research, data-curation and ML-platform for a fleet of autonomous vehicles. Before that Oscar was a postdoctoral scholar in Trevor Darrell’s lab at Berkeley AI Research where he worked on automated quantification of scientific image-data using deep learning.

Previously, Oscar was lead developer at Hovding where he created the algorithmic framework and hardware design for their invisible bicycle helmet. Oscar has also worked on automated dietary logging systems for consumer applications and focusing algorithms for image-based cell analysis.

Oscar Beijbom manages and develops CoralNet, deploying deep convolutional neural networks to help coral reef ecologists mine image data.

Oscar Beijom got his Ph.D in computer vision and machine learning at UCSD under David Kriegman and Serge Belongie, and his M.S. in engineering physics at Lund University under Kalle Astrom.

Hosted by Professor Stefano Soatto

Via: Zoom Webinar