Safety Metrics for Semantic Segmentation in Autonomous Driving
arXiv:2105.10142
Abstract
Within the context of autonomous driving, safety-related metrics for deep neural networks have been widely studied for image classification and object detection. In this paper, we further consider safety-aware correctness and robustness metrics specialized for semantic segmentation. The novelty of our proposal is to move beyond pixel-level metrics: Given two images with each having N pixels being class-flipped, the designed metrics should, depending on the clustering of pixels being class-flipped or the location of occurrence, reflect a different level of safety criticality. The result evaluated on an autonomous driving dataset demonstrates the validity and practicality of our proposed methodology.
arXiv commentPaper accepted at IEEE AI Test'21
References in corpus (6)
- MobileNetV2: Inverted Residuals and Linear Bottlenecks
- ENet: A Deep Neural Network Architecture for Real-Time Semantic Segmentation
- The Cityscapes Dataset for Semantic Urban Scene Understanding
- A Safety Framework for Critical Systems Utilising Deep Neural Networks
- A Framework for Assurance of Medication Safety using Machine Learning
- Towards the Quantification of Safety Risks in Deep Neural Networks