Ramifications of incorrect image segmentations; emphasizing on the potential effects on deep learning methods failure

Ramifications of incorrect image segmentations; emphasizing on the potential effects on deep learning methods failure

Abstract Introduction Detecting failure cases is critical to ensure a secure self-driving system. Any flaw in the system directly results in an accident. In genuine class, the model’s probability reflects better-reflected model confidence. As a result, the confidence distributions of failed predicti...

Full description

Bibliographic Details
Main Author: Hayat Al-Dmour
Format: Article
Language:English
Published: SpringerOpen 2022-05-01
Series:Journal of Big Data
Subjects:
Deep learning model
Failure detection
Image classification
Image segmentation
Neural network
Ramifications
Online Access:https://doi.org/10.1186/s40537-022-00624-0

Internet

https://doi.org/10.1186/s40537-022-00624-0

Similar Items