A Fine-Grained Dataset and its Efficient Semantic Segmentation for Unstructured Driving Scenarios
This addresses the problem of autonomous driving in challenging unstructured outdoor environments for researchers, though it is incremental as it builds on existing semantic segmentation methods.
The authors tackled the lack of labeled datasets for autonomous driving in unstructured environments by introducing TAS500, a fine-grained semantic segmentation dataset, which improved overall prediction accuracy, especially along class boundaries.
Research in autonomous driving for unstructured environments suffers from a lack of semantically labeled datasets compared to its urban counterpart. Urban and unstructured outdoor environments are challenging due to the varying lighting and weather conditions during a day and across seasons. In this paper, we introduce TAS500, a novel semantic segmentation dataset for autonomous driving in unstructured environments. TAS500 offers fine-grained vegetation and terrain classes to learn drivable surfaces and natural obstacles in outdoor scenes effectively. We evaluate the performance of modern semantic segmentation models with an additional focus on their efficiency. Our experiments demonstrate the advantages of fine-grained semantic classes to improve the overall prediction accuracy, especially along the class boundaries. The dataset and pretrained model are available at mucar3.de/icpr2020-tas500.