A Little Fog for a Large Turn

Abstract

Small, carefully crafted perturbations called adversarial perturbations can easily fool neural networks. However, these perturbations are largely additive and not naturally found. We turn our attention to the field of Autonomous navigation wherein adverse weather conditions such as fog have a drastic effect on the predictions of these systems. These weather conditions are capable of acting like natural adversaries that can help in testing models. To this end, we introduce a general notion of adversarial perturbations, which can be created using generative models and provide a methodology inspired by Cycle-Consistent Generative Adversarial Networks to generate adversarial weather conditions for a given image. Our formulation and results show that these images provide a suitable testbed for steering models used in Autonomous navigation models. Our work also presents a more natural and general definition of Adversarial perturbations based on Perceptual Similarity.

arXiv commentAccepted to WACV 2020

References in corpus (12)

• Explaining and Harnessing Adversarial Examples
• End to End Learning for Self-Driving Cars
• Semantic Foggy Scene Understanding with Synthetic Data
• Robust Physical-World Attacks on Deep Learning Models
• CARLA: An Open Urban Driving Simulator
• DeepTest: Automated Testing of Deep-Neural-Network-driven Autonomous Cars
• Natural Adversarial Examples
• Learning a Driving Simulator
• Why do deep convolutional networks generalize so poorly to small image transformations?
• Computer Vision for Autonomous Vehicles: Problems, Datasets and State of the Art
• Adversarial Patch
• DeepRoad: GAN-based Metamorphic Autonomous Driving System Testing

Cited by in corpus (2)

• Multimodal End-to-End Learning for Autonomous Steering in Adverse Road and Weather Conditions
• RiWNet: A moving object instance segmentation Network being Robust in adverse Weather conditions