Multi-Label Continual Learning for the Medical Domain: A Novel Benchmark

arXiv:2404.06859v1 Announce Type: new
Abstract: Multi-label image classification in dynamic environments is a problem that poses significant challenges. Previous studies have primarily focused on scenarios such as Domain Incremental Learning and Class Incremental Learning, which do not fully capture the complexity of real-world applications. In this paper, we study the problem of classification of medical imaging in the scenario termed New Instances \& New Classes, which combines the challenges of both new class arrivals and domain shifts in a single …

abstract applications arxiv benchmark challenges class classification complexity continual cs.ai cs.cv domain dynamic environments image incremental medical novel paper studies study type world