ePrints@IIScePrints@IISc Home | About | Browse | Latest Additions | Advanced Search | Contact | Help

Towards Efficient and Effective Self-supervised Learning of Visual Representations

Addepalli, S and Bhogale, K and Dey, P and Babu, RV (2022) Towards Efficient and Effective Self-supervised Learning of Visual Representations. In: 17th European Conference on Computer Vision, ECCV 2022, 23 - 27 October 2022, Tel Aviv, pp. 523-538.

ECCV_2022.pdf - Published Version

Download (1MB) | Preview
Official URL: https://doi.org/10.1007/978-3-031-19821-2_30


Self-supervision has emerged as a propitious method for visual representation learning after the recent paradigm shift from handcrafted pretext tasks to instance-similarity based approaches. Most state-of-the-art methods enforce similarity between various augmentations of a given image, while some methods additionally use contrastive approaches to explicitly ensure diverse representations. While these approaches have indeed shown promising direction, they require a significantly larger number of training iterations when compared to the supervised counterparts. In this work, we explore reasons for the slow convergence of these methods, and further propose to strengthen them using well-posed auxiliary tasks that converge significantly faster, and are also useful for representation learning. The proposed method utilizes the task of rotation prediction to improve the efficiency of existing state-of-the-art methods. We demonstrate significant gains in performance using the proposed method on multiple datasets, specifically for lower training epochs. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Item Type: Conference Paper
Publication: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Publisher: Springer Science and Business Media Deutschland GmbH
Additional Information: The copyright for this article belongs to Springer Science and Business Media Deutschland GmbH.
Keywords: Multiple data sets; Paradigm shifts; Performance; Slow convergences; State-of-the-art methods; Training epochs; Visual representations, Learning systems
Department/Centre: Division of Interdisciplinary Sciences > Computational and Data Sciences
Date Deposited: 09 Jan 2023 07:31
Last Modified: 09 Jan 2023 07:31
URI: https://eprints.iisc.ac.in/id/eprint/78924

Actions (login required)

View Item View Item