Piecewise classifier mappings: Learning fine-grained learners for novel categories with few examples

TL;DR

This paper introduces a novel piecewise classifier mapping approach for few-shot fine-grained recognition, enabling deep models to classify new categories with very few examples by learning a set of sub-classifiers.

Contribution

The paper proposes a new end-to-end deep network with piecewise mappings for classifier generation, tailored for few-shot fine-grained recognition, improving generalization to novel categories.

Findings

Achieves superior accuracy on three fine-grained datasets.

Effectively generalizes to novel categories with fewer than five examples.

Outperforms existing few-shot learning baselines.

Abstract

Humans are capable of learning a new fine-grained concept with very little supervision, \emph{e.g.}, few exemplary images for a species of bird, yet our best deep learning systems need hundreds or thousands of labeled examples. In this paper, we try to reduce this gap by studying the fine-grained image recognition problem in a challenging few-shot learning setting, termed few-shot fine-grained recognition (FSFG). The task of FSFG requires the learning systems to build classifiers for novel fine-grained categories from few examples (only one or less than five). To solve this problem, we propose an end-to-end trainable deep network which is inspired by the state-of-the-art fine-grained recognition model and is tailored for the FSFG task. Specifically, our network consists of a bilinear feature learning module and a classifier mapping module: while the former encodes the discriminative information of an exemplar image into a feature vector, the latter maps the intermediate feature into the decision boundary of the novel category. The key novelty of our model is a "piecewise mappings" function in the classifier mapping module, which generates the decision boundary via learning a set of more attainable sub-classifiers in a more parameter-economic way. We learn the exemplar-to-classifier mapping based on an auxiliary dataset in a meta-learning fashion, which is expected to be able to generalize to novel categories. By conducting comprehensive experiments on three fine-grained datasets, we demonstrate that the proposed method achieves superior performance over the competing baselines.