ISD: Self-Supervised Learning by Iterative Similarity Distillation

Tejankar, Ajinkya; Koohpayegani, Soroush Abbasi; Pillai, Vipin; Favaro, Paolo; Pirsiavash, Hamed (October 2021). ISD: Self-Supervised Learning by Iterative Similarity Distillation. International Conference on Computer Vision, pp. 9589-9598. IEEE 10.1109/ICCV48922.2021.00947

	Text Conference_ISD_Self-Supervised_Learning_by_Iterative_Similarity_Distillation.pdf - Accepted Version Restricted to registered users only Available under License Publisher holds Copyright. Download (2MB)
	Text ISD_Self-Supervised_Learning_by_Iterative_Similarity_Distillation.pdf - Published Version Restricted to registered users only Available under License Publisher holds Copyright. Download (3MB)

Recently, contrastive learning has achieved great results in self-supervised learning, where the main idea is to pull two augmentations of an image (positive pairs) closer compared to other random images (negative pairs). We argue that not all negative images are equally negative. Hence, we introduce a self-supervised learning algorithm where we use a soft similarity for the negative images rather than a binary distinction between positive and negative pairs. We iteratively distill a slowly evolving teacher model to the student model by capturing the similarity of a query image to some random images and transferring that knowledge to the student. Specifically, our method should handle unbalanced and unlabeled data better than existing contrastive learning methods, because the randomly chosen negative set might include many samples that are semantically similar to the query image. In this case, our method labels them as highly similar while standard contrastive methods label them as negatives. Our method achieves comparable results to the state-of-the-art models.

Interest & Impact

Downloads

2 since deposited on 21 Apr 2022
0 in the past 12 months

Citations

5 Citations in Web of Science ®
6 Citations in Scopus

Search

in Google Scholar™

Services

Actions (login required)

Edit item

Actions (login required)

Edit item