Transient Fault Tolerant Semantic Segmentation for Autonomous Driving - Archive ouverte HAL
Communication Dans Un Congrès Année : 2024

Transient Fault Tolerant Semantic Segmentation for Autonomous Driving

Résumé

Deep learning models are crucial for autonomous vehicle perception, but their reliability is challenged by algorithmic limitations and hardware faults. We address the latter by examining fault-tolerance in semantic segmentation models. Using established hardware fault models, we evaluate existing hardening techniques both in terms of accuracy and uncertainty and introduce ReLUMax, a novel simple activation function designed to enhance resilience against transient faults. ReLUMax integrates seamlessly into existing architectures without time overhead. Our experiments demonstrate that ReLUMax effectively improves robustness, preserving performance and boosting prediction confidence, thus contributing to the development of reliable autonomous driving systems. Code available at: https://github.com/iurada/neutronsegmentation.
Fichier principal
Vignette du fichier
NeutronSegm_ExtendedAbstract_ECCV2024_camera_ready.pdf (5.07 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04684784 , version 1 (03-09-2024)

Licence

Identifiants

  • HAL Id : hal-04684784 , version 1

Citer

Leonardo Iurada, Niccolò Cavagnero, Fernando Fernandes dos Santos, Giuseppe Averta, Paolo Rech, et al.. Transient Fault Tolerant Semantic Segmentation for Autonomous Driving. UNCV 2024 - 3rd Workshop on Uncertainty Quantification for Computer Vision, Sep 2024, Milano, Italy. pp.1-6. ⟨hal-04684784⟩
8 Consultations
5 Téléchargements

Partager

More