Alpha Net: Adaptation with Composition in Classifier Space

Nadine Chang, Jayanth Koushik, Michael Tarr, Martial Hebert, Yu-Xiong Wang

Keywords: long-tail recognition, classifier composition

Abstract: Deep learning classification models typically train poorly on classes with small numbers of examples. Motivated by the human ability to solve this task, models have been developed that transfer knowledge from classes with many examples to learn classes with few examples. Critically, the majority of these models transfer knowledge within model feature space. In this work, we demonstrate that transferring knowledge within classifier space is more effective and efficient. Specifically, by linearly combining strong nearest neighbor classifiers along with a weak classifier, we are able to compose a stronger classifier. Uniquely, our model can be implemented on top of any existing classification model that includes a classifier layer. We showcase the success of our approach in the task of long-tailed recognition, whereby the classes with few examples, otherwise known as the tail classes, suffer the most in performance and are the most challenging classes to learn. Using classifier-level knowledge transfer, we are able to drastically improve - by a margin as high as 10.5% - the state-of-the-art performance on the tail categories.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

One-sentence Summary: In this work, we demonstrate that transferring knowledge within classifier space can greatly improve the recognition performance on rare categories in the long-tail.

Community Implementations: [ 4 code implementations](https://www.catalyzex.com/paper/arxiv:2008.07073/code)

Reviewed Version (pdf): https://openreview.net/references/pdf?id=44QI4ykscR