Michal Uřičář
Advanced Dataset Augmentation
Anotace: Autonomous driving is a specific area of computer vision and machine learning with a focus on safety. Data acquisition is a very tedious and costly task and, it is not possible to acquire data containing all admissible corner cases. Thanks to the recent progress in generative modeling it is possible to tackle this problem via advanced dataset augmentation, where the missing data are created artificially and used for both training and testing of the classifiers used in the autonomous driving challenges. We present our recent work on this topic for Soiling and Adverse Weather Classification (SAW) and Pedestrian Detection.

Ondřej Zeman
Use of simulation and virtual reality in car development
Anotace: The future of validation in automotive lies in the use of virtual environment and tools that allow companies to move from real world testing to simulated environment. The presentation will showcase one of the tools, with practical demonstration on how sensor simulation can be utilized as a virtual validation method, to extend the methods for ADAS system validation.

Jan Olšina
Ground Truth Extraction from real world data
Anotace: The process of ground truth labeling from a set of reference measurements is inevitable, yet demanding and expensive task within the development of highly automated and autonomous driving (HAD/AD) systems. The presentation will illustrate an effective annotation process that enables automation on different levels. The presented approach is based on profound know-how of the challenges the annotation process brings when used for statistical validation of HAD/AD systems combined with the utilization of the state of art machine learning techniques.