How the perception system of Yandex's delivery robot works
summary
Describes Yandex's autonomous delivery robot perception system: how cameras, sensors and perception software detect pedestrians, vehicles and unexpected obstacles to enable safe navigation in real urban environment. Explains the system architecture and key perception technologies used to recognize and avoid moving and static hazards.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 1:36Problem and requirements
- 4:51Robot sensors
- 7:20Perception models
- 10:50Base perception model architecture
- 13:45Traffic light detection model
- 16:10Model inference and optimization
- 18:40System testing
- 22:32Conclusion and future work
- 24:39Q&A
related talks
27mDriving Autonomous Vehicles Forward w/Real World Applications
09mDelivery Robots
49mUsing Deep Learning for Perception in Autonomous Systems: A Perspective from Control Theory
13mEnhancing Object Detection & Avoidance Systems for Robots in Challenging Environments Using AI
04mAI and the Robotics Revolution
26mAutonomy – The Era of Vehicles Without Drivers
11mInterview with Ahti Heinla
29mHow Amazon builds AI-enabled perception for robots