loading
Papers Papers/2022 Papers Papers/2022

Research.Publish.Connect.

Paper

Authors: Qin Wang 1 ; Kai Krajsek 2 and Hanno Scharr 1

Affiliations: 1 IAS-8: Data Analytics and Machine Learning, Forschungszentrum Jülich, Germany ; 2 Jülich Supercomputing Centre (JSC), Forschungszentrum Jülich, Germany

Keyword(s): Self-Supervised Learning, Augmentation, Vision Transformer.

Abstract: Many recent self-supervised pretraining methods use augmented versions of the same image as samples for their learning schemes. We observe that ’easy’ samples, i.e. samples being too similar to each other after augmentation, have only limited value as learning signal. We therefore propose to rescue easy samples and make them harder. To do so, we select the top k easiest samples using cosine similarity, strongly augment them, forward-pass them through the model, calculate cosine similarity of the output as loss, and add it to the original loss in a weighted fashion. This method can be adopted to all contrastive or other augmented-pair based learning methods, whether they involve negative pairs or not, as it changes handling of easy positives, only. This simple but effective approach introduces greater variability into such self-supervised pretraining processes, significantly increasing the performance on various downstream tasks as observed in our experiments. We pretrain models of di fferent sizes, i.e. ResNet-50, ViT-S, ViT-B, or ViT-L, using ImageNet with SimCLR, MoCo v3, or DINOv2 training schemes. Here, e.g., we consistently find to improve results for ImageNet top-1 accuracy with a linear classifier establishing new SOTA for this task. (More)

CC BY-NC-ND 4.0

Sign In Guest: Register as new SciTePress user now for free.

Sign In SciTePress user: please login.

PDF ImageMy Papers

You are not signed in, therefore limits apply to your IP address 3.144.89.220

In the current month:
Recent papers: 100 available of 100 total
2+ years older papers: 200 available of 200 total

Paper citation in several formats:
Wang, Q., Krajsek, K. and Scharr, H. (2025). Rescuing Easy Samples in Self-Supervised Pretraining. In Proceedings of the 20th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 2: VISAPP; ISBN 978-989-758-728-3; ISSN 2184-4321, SciTePress, pages 400-409. DOI: 10.5220/0013167900003912

@conference{visapp25,
author={Qin Wang and Kai Krajsek and Hanno Scharr},
title={Rescuing Easy Samples in Self-Supervised Pretraining},
booktitle={Proceedings of the 20th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 2: VISAPP},
year={2025},
pages={400-409},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013167900003912},
isbn={978-989-758-728-3},
issn={2184-4321},
}

TY - CONF

JO - Proceedings of the 20th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 2: VISAPP
TI - Rescuing Easy Samples in Self-Supervised Pretraining
SN - 978-989-758-728-3
IS - 2184-4321
AU - Wang, Q.
AU - Krajsek, K.
AU - Scharr, H.
PY - 2025
SP - 400
EP - 409
DO - 10.5220/0013167900003912
PB - SciTePress