On the Effectiveness of Supervision in Asymmetric Non-Contrastive Learning

Jeongheon Oh, Kibok Lee
Proceedings of the 41st International Conference on Machine Learning, PMLR 235:38541-38561, 2024.

Abstract

Supervised contrastive representation learning has been shown to be effective in various transfer learning scenarios. However, while asymmetric non-contrastive learning (ANCL) often outperforms its contrastive learning counterpart in self-supervised representation learning, the extension of ANCL to supervised scenarios is less explored. To bridge the gap, we study ANCL for supervised representation learning, coined SupSiam and SupBYOL, leveraging labels in ANCL to achieve better representations. The proposed supervised ANCL framework improves representation learning while avoiding collapse. Our analysis reveals that providing supervision to ANCL reduces intra-class variance, and the contribution of supervision should be adjusted to achieve the best performance. Experiments demonstrate the superiority of supervised ANCL across various datasets and tasks. The code is available at: https://github.com/JH-Oh-23/Sup-ANCL.

Cite this Paper

BibTeX
@InProceedings{pmlr-v235-oh24a, title = {On the Effectiveness of Supervision in Asymmetric Non-Contrastive Learning}, author = {Oh, Jeongheon and Lee, Kibok}, booktitle = {Proceedings of the 41st International Conference on Machine Learning}, pages = {38541--38561}, year = {2024}, editor = {Salakhutdinov, Ruslan and Kolter, Zico and Heller, Katherine and Weller, Adrian and Oliver, Nuria and Scarlett, Jonathan and Berkenkamp, Felix}, volume = {235}, series = {Proceedings of Machine Learning Research}, month = {21--27 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v235/main/assets/oh24a/oh24a.pdf}, url = {https://proceedings.mlr.press/v235/oh24a.html}, abstract = {Supervised contrastive representation learning has been shown to be effective in various transfer learning scenarios. However, while asymmetric non-contrastive learning (ANCL) often outperforms its contrastive learning counterpart in self-supervised representation learning, the extension of ANCL to supervised scenarios is less explored. To bridge the gap, we study ANCL for supervised representation learning, coined SupSiam and SupBYOL, leveraging labels in ANCL to achieve better representations. The proposed supervised ANCL framework improves representation learning while avoiding collapse. Our analysis reveals that providing supervision to ANCL reduces intra-class variance, and the contribution of supervision should be adjusted to achieve the best performance. Experiments demonstrate the superiority of supervised ANCL across various datasets and tasks. The code is available at: https://github.com/JH-Oh-23/Sup-ANCL.} }
Endnote
%0 Conference Paper %T On the Effectiveness of Supervision in Asymmetric Non-Contrastive Learning %A Jeongheon Oh %A Kibok Lee %B Proceedings of the 41st International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2024 %E Ruslan Salakhutdinov %E Zico Kolter %E Katherine Heller %E Adrian Weller %E Nuria Oliver %E Jonathan Scarlett %E Felix Berkenkamp %F pmlr-v235-oh24a %I PMLR %P 38541--38561 %U https://proceedings.mlr.press/v235/oh24a.html %V 235 %X Supervised contrastive representation learning has been shown to be effective in various transfer learning scenarios. However, while asymmetric non-contrastive learning (ANCL) often outperforms its contrastive learning counterpart in self-supervised representation learning, the extension of ANCL to supervised scenarios is less explored. To bridge the gap, we study ANCL for supervised representation learning, coined SupSiam and SupBYOL, leveraging labels in ANCL to achieve better representations. The proposed supervised ANCL framework improves representation learning while avoiding collapse. Our analysis reveals that providing supervision to ANCL reduces intra-class variance, and the contribution of supervision should be adjusted to achieve the best performance. Experiments demonstrate the superiority of supervised ANCL across various datasets and tasks. The code is available at: https://github.com/JH-Oh-23/Sup-ANCL.
APA
Oh, J. & Lee, K.. (2024). On the Effectiveness of Supervision in Asymmetric Non-Contrastive Learning. Proceedings of the 41st International Conference on Machine Learning, in Proceedings of Machine Learning Research 235:38541-38561 Available from https://proceedings.mlr.press/v235/oh24a.html.

Related Material