Combining clustering and active learning for the detection and learning of new image classes
Nenhuma Miniatura disponível
Data
2019-09-17
Autores
Coletta, Luiz F. S. [UNESP]
Ponti, Moacir
Hruschka, Eduardo R.
Acharya, Ayan
Ghosh, Joydeep
Título da Revista
ISSN da Revista
Título de Volume
Editor
Elsevier B.V.
Resumo
Discriminative classification models often assume all classes are available at the training phase. As such models do not have a strategy to learn new concepts from available unlabeled instances, they usually work poorly when unknown classes emerge from future data to be classified. To address the appearance of new classes, some authors have developed approaches to transfer knowledge from known to unknown classes. Our study provides a more flexible approach to learn new (visual) classes that emerge over time. The key idea is materialized by an iterative classifier that combines Support Vector Machines with clustering via an optimization algorithm. An entropy and density-based selection strategy explores label uncertainty and high-density regions from unlabeled data to be classified. Selected instances from new classes are submitted to get labels and then used to improve the model. The proposed image classifier is consistently better than approaches that select instances randomly or from clusters. We also show that features obtained via Deep Learning methods improve results when compared with shallow features, but only using our selection strategy. Our approach requires fewer iterations to learn new classes, thereby significantly saving labeling costs. (C) 2019 Elsevier B.V. All rights reserved.
Descrição
Palavras-chave
Image classification, Active learning, Clustering, Open set, Deep learning
Como citar
Neurocomputing. Amsterdam: Elsevier, v. 358, p. 150-165, 2019.