Marco Toldo
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Marco Toldo has a total of 3 co-authors. Here are the top co-authors by number of shared publications:
- Francesco Barbato
Department of Information Engineering, University of Padova, Padua, Italy
1 shared publication - Pietro Zanuttigh
University of Padova
1 shared publication - Umberto Michieli
University of Padua
1 shared publication
Latest Publications
- Open Access
- 15 April 2023
- Published by Springer Nature in The Visual Computer
- Vol. 40 (2) , 811-830
Road scenes segmentation across different domains by disentangling latent representations
Deep learning models obtain impressive accuracy in road scene understanding; however, they need a large number of labeled samples for their training. Additionally, such models do not generalize well to environments where the statistical properties of data do not perfectly match those of training scenes, and this can be a significant problem for intelligent vehicles. Hence, domain adaptation approaches have been introduced to transfer knowledge acquired on a label-abundant source domain to a related label-scarce target domain. In this work, we design and carefully analyze multiple latent space-shaping regularization strategies that work together to reduce the domain shift. More in detail, we devise a feature clustering strategy to increase domain alignment, a feature perpendicularity constraint to space apart features belonging to different semantic classes, including those not present in the current batch, and a feature norm alignment strategy to separate active and inactive channels. In addition, we propose a novel evaluation metric to capture the relative performance of an adapted model with respect to supervised training. We validate our framework in driving scenarios, considering both synthetic-to-real and real-to-real adaptation, outperforming previous feature-level state-of-the-art methods on multiple road scenes benchmarks.
- 15 April 2023
- Published by Springer Nature in The Visual Computer
- Vol. 40 (2) , 811-830
Open AccessRoad scenes segmentation across different domains by disentangling latent representations
Deep learning models obtain impressive accuracy in road scene understanding; however, they need a large number of labeled samples for their training. Additionally, such models do not generalize well to environments where the statistical properties of data do not perfectly match those of training scenes, and this can be a significant problem for intelligent vehicles. Hence, domain adaptation approaches have been introduced to transfer knowledge acquired on a label-abundant source domain to a related label-scarce target domain. In this work, we design and carefully analyze multiple latent space-shaping regularization strategies that work together to reduce the domain shift. More in detail, we devise a feature clustering strategy to increase domain alignment, a feature perpendicularity constraint to space apart features belonging to different semantic classes, including those not present in the current batch, and a feature norm alignment strategy to separate active and inactive channels. In addition, we propose a novel evaluation metric to capture the relative performance of an adapted model with respect to supervised training. We validate our framework in driving scenarios, considering both synthetic-to-real and real-to-real adaptation, outperforming previous feature-level state-of-the-art methods on multiple road scenes benchmarks.