Sam, DB and Agarwalla, A and Radhakrishnan, VB (2022) Beyond Learning Features: Training a Fully-Functional Classifier with ZERO Instance-Level Labels. In: 36th AAAI Conference on Artificial Intelligence, AAAI 2022, 22 February - 1 March 2022, Virtual, Online, pp. 2162-2170.
PDF
AAAI_2022.pdf - Published Version Restricted to Registered users only Download (4MB) | Request a copy |
Abstract
We attempt to train deep neural networks for classification without using any labeled data. Existing unsupervised methods, though mine useful clusters or features, require some annotated samples to facilitate the final task-specific predictions. This defeats the true purpose of unsupervised learning and hence we envisage a paradigm of 'true' self-supervision, where absolutely no annotated instances are used for training a classifier. The proposed method first pretrains a deep network through self-supervision and performs clustering on the learned features. A classifier layer is then appended to the self-supervised network and is trained by matching the distribution of the predictions to that of a predefined prior. This approach leverages the distribution of labels for supervisory signals and consequently, no image-label pair is needed. Experiments reveal that the method works on major nominal as well as ordinal classification datasets and delivers significant performance. Copyright © 2022, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.
Item Type: | Conference Paper |
---|---|
Publication: | Proceedings of the 36th AAAI Conference on Artificial Intelligence, AAAI 2022 |
Publisher: | Association for the Advancement of Artificial Intelligence |
Additional Information: | The copyright for this article belongs to Association for the Advancement of Artificial Intelligence. |
Keywords: | Classification (of information), Classification datasets; Clusterings; Labeled data; Matchings; Ordinal classification; Performance; Supervised network; Unsupervised method, Deep neural networks |
Department/Centre: | Division of Interdisciplinary Sciences > Computational and Data Sciences |
Date Deposited: | 21 Feb 2023 04:58 |
Last Modified: | 21 Feb 2023 04:58 |
URI: | https://eprints.iisc.ac.in/id/eprint/80575 |
Actions (login required)
View Item |