Street-Map Based Validation of Semantic Segmentation in Autonomous Driving
This addresses the need for cost-effective validation of AI models in autonomous driving, though it is incremental as it builds on existing validation methods by incorporating map data.
The paper tackles the problem of validating semantic segmentation models for autonomous driving by proposing a model-agnostic approach using street maps to avoid costly ground truth data, and demonstrates its potential on the Cityscapes dataset by uncovering errors in segmentation masks.
Artificial intelligence for autonomous driving must meet strict requirements on safety and robustness, which motivates the thorough validation of learned models. However, current validation approaches mostly require ground truth data and are thus both cost-intensive and limited in their applicability. We propose to overcome these limitations by a model agnostic validation using a-priori knowledge from street maps. In particular, we show how to validate semantic segmentation masks and demonstrate the potential of our approach using OpenStreetMap. We introduce validation metrics that indicate false positive or negative road segments. Besides the validation approach, we present a method to correct the vehicle's GPS position so that a more accurate localization can be used for the street-map based validation. Lastly, we present quantitative results on the Cityscapes dataset indicating that our validation approach can indeed uncover errors in semantic segmentation masks.