Special Offer: Get 50% off your first 2 months when you do one of the following
Personalized offer codes will be given in each session
About This Webinar

Autonomous vehicles (AVs) are expected to dramatically redefine the future of transportation. A key challenge researchers are solving is how to create models that are robust and reliable enough to predict the motion of traffic agents such as cars, cyclists, and pedestrians. These models are key in planning an AV's decisions. Even though there has been a surge of interest in this problem, there still isn't a clear winning technology for future motion prediction that satisfies all of the needs of AVs.

Learn how Lyft Level 5 is thinking about applying deep prediction and machine learning frameworks to create motion prediction models, and explore the challenges involved in building self-driving systems. After this webinar, you'll be able to apply what you've learned to our Prediction Dataset—the largest released to date.

Agenda
  • About Lyft Level 5
  • Deep Prediction and Autonomous Vehicles
  • The Level 5 Motion Prediction Competition
  • Q&A
When: Tuesday, July 28, 2020 · 09:00:00 AM · Pacific Time (US & Canada)
Duration: 1 hour 30 minutes
Language: English
Who can attend? Anyone with the event link can attend
Dial-in available? (listen only): No
Featured Presenters
Webinar hosting presenter
Director of Advanced Autonomous Vehicle Development
Prior to leading Advance AV Development at Lyft Level 5, Sacha was the Director of Engineering at Waymo. He spent seven years there leading engineering teams across various initiatives to organize and data mine all of Google’s geo imagery, including developing some of the largest crowd-sourcing solutions and deep-net based computer vision to contribute to Google Maps data.
Webinar hosting presenter
Head of Autonomous Vehicles Research
At Lyft Level 5, Peter leads a machine learning team focused on novel approaches to perception, prediction, planning, and simulation. Previously, he co-founded and led Blue Vision Labs, which developed city-scale crowd-sourced SLAM for use in augmented reality and robotics that was acquired by Lyft in 2018. Peter has a PhD in Robotics from Oxford Robotics Institute and has dozens of patents and publications in the field of machine learning, computer vision and its applications to robotics.
Webinar hosting presenter
Staff Engineer
John Houston has over twelve years of experience with applied autonomous vehicles. He designed the initial on-car autonomy software platform which includes real-time messaging, service infrastructure, logging, replay, configuration, and visualization systems. As the tech lead of the prediction team, he pioneered Lyft's machine-learned obstacle motion prediction system called Deep Prediction. John holds a Masters of Science in Computer Science from Tufts University.
Webinar hosting presenter
Software Engineer
Vladimir focuses on computer vision at Level 5. He has extensive experience participating and winning various machine learning competitions at Kaggle, CVPR, and MICCAI and holds the title of Kaggle Grandmaster. In his free time, he's developing an open-source image augmentation library called Albumentations, which was used in all winning solutions in the computer vision challenges at Kaggle in the last year. He has published more than 16 journal and conference papers.
Hosted By
Lyft webinar platform hosts Motion Prediction for Self-Driving Cars
Lyft Level 5 is developing a full self-driving system for the Lyft network and democratizing access to this technology for millions of riders. Learn more at lyft.com/level5.