The Performance of Transferability Metrics Does Not Translate to Medical Tasks

Autor(es)

Chaves, Levy ; Bissoto, Alceu ; Valle, Eduardo ; Avila, Sandra

• Tipo de documento: Artigo Científico
• Fonte: DOMAIN ADAPTATION AND REPRESENTATION TRANSFER, DART 2023; v. 14293, p. 10-pg., 2024-01-01.

Resumo

Transfer learning boosts the performance of medical image analysis by enabling deep learning (DL) on small datasets through the knowledge acquired from large ones. As the number of DL architectures explodes, exhaustively attempting all candidates becomes unfeasible, motivating cheaper alternatives for choosing them. Transferability scoring methods emerge as an enticing solution, allowing to efficiently calculate a score that correlates with the architecture accuracy on any target dataset. However, since transferability scores have not been evaluated on medical datasets, their use in this context remains uncertain, preventing them from benefiting practitioners. We fill that gap in this work, thoroughly evaluating seven transferability scores in three medical applications, including out-of-distribution scenarios. Despite promising results in general-purpose datasets, our results show that no transferability score can reliably and consistently estimate target performance in medical contexts, inviting further work in that direction. (AU)

• Processo FAPESP: 13/08293-7 - CECC - Centro de Engenharia e Ciências Computacionais
• Beneficiário: Munir Salomao Skaf
• Modalidade de apoio: Auxílio à Pesquisa - Centros de Pesquisa, Inovação e Difusão - CEPIDs
• Processo FAPESP: 20/09838-0 - BI0S - Brazilian Institute of Data Science
• Beneficiário: João Marcos Travassos Romano
• Modalidade de apoio: Auxílio à Pesquisa - Programa Centros de Pesquisa em Engenharia
• Processo FAPESP: 22/09606-8 - Compreensão do papel de atalhos e mudanças de distribuição para generalização de redes neurais profundas
• Beneficiário: Alceu Emanuel Bissoto
• Modalidade de apoio: Bolsas no Exterior - Estágio de Pesquisa - Doutorado
• Processo FAPESP: 19/19619-7 - Geração ilimitada de imagens de lesões de pele usando redes generativas adversariais
• Beneficiário: Alceu Emanuel Bissoto
• Modalidade de apoio: Bolsas no Brasil - Doutorado