A comparative study of calibration methods for imbalanced class incremental learning

• Published in: Multimedia tools and applications Vol. 81; no. 14; pp. 19237 - 19256
• Main Authors: Aggarwal, Umang, Popescu, Adrian, Belouadah, Eden, Hudelot, Celine
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.06.2022; Springer Nature B.V; Springer Verlag
• Subjects: 1182: Deep Processing of Multimedia Data; Algorithms; Artificial Intelligence; Calibration; Comparative studies; Computer Communication Networks; Computer Science; Computer Vision and Pattern Recognition; Data Structures and Information Theory; Datasets; Deep learning; Distillation; Information management; Machine learning; Multimedia Information Systems; Special Purpose and Application-Based Systems; Visual tasks; Incremental learning; Calibration; Class imbalance; Image classification

Deep learning approaches are successful in a wide range of AI problems and in particular for visual recognition tasks. However, there are still open problems among which is the capacity to handle streams of visual information and the management of class imbalance in datasets. Existing research appro...