Testing DNN-based Autonomous Driving Systems under
               Critical Environmental Conditions

                Zhong Li 1 2 Minxue Pan 1 3 Tian Zhang 1 2 Xuandong Li 1 2

             Abstract
   Due to the increasing usage of Deep Neural Net-
   work (DNN) based autonomous driving systems
   (ADS) where erroneous or unexpected behaviours
   can lead to catastrophic accidents, testing such
   systems is of growing importance. Existing ap-
   proaches often just focus on finding erroneous
  ...