Few-shot learning with simplex

Anonymous

Nov 07, 2017 (modified: Nov 07, 2017) ICLR 2018 Conference Blind Submission readers: everyone Show Bibtex
  • Abstract: Deep learning has made remarkable achievement in many fields. However, learning parameters of a neural networks usually needs a large amount of labeled data. The algorithms of deep learning, therefore, encounter difficulty when applied to supervised learning where only little data are available. This problem is called one-shot learning. To address it, we propose a novel algorithm for few-shot learning using discrete geometry, in the sense that the samples in a class are modeled as a reduced simplex. The volume of the simplex is used for the measurement of class scatter. During testing, combined with the test sample and the points in the class, a new simplex is formed. Then the similarity between the test sample and the class can be quantized with the ratio of volumes of the new simplex to the original class simplex. Moreover, we present an approach to constructing simplices using local regions of feature maps yielded by convolutional neural networks. Experiments on Omniglot and miniImageNet verify the superiority of our simplex algorithm on few-shot learning.
  • TL;DR: A simplex-based geometric method is proposed to cope with few-shot learning problems.
  • Keywords: one-shot learning, few-shot learning, deep learning, simplex

Loading