LeOCLR: Leveraging Original Images for Contrastive Learning of Visual Representations
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909583577972736 |
|---|---|
| author | Alkhalefi, Mohammad Leontidis, Georgios Zhong, Mingjun |
| author_facet | Alkhalefi, Mohammad Leontidis, Georgios Zhong, Mingjun |
| contents | Contrastive instance discrimination methods outperform supervised learning in downstream tasks such as image classification and object detection. However, these methods rely heavily on data augmentation during representation learning, which can lead to suboptimal results if not implemented carefully. A common augmentation technique in contrastive learning is random cropping followed by resizing. This can degrade the quality of representation learning when the two random crops contain distinct semantic content. To tackle this issue, we introduce LeOCLR (Leveraging Original Images for Contrastive Learning of Visual Representations), a framework that employs a novel instance discrimination approach and an adapted loss function. This method prevents the loss of important semantic features caused by mapping different object parts during representation learning. Our experiments demonstrate that LeOCLR consistently improves representation learning across various datasets, outperforming baseline models. For instance, LeOCLR surpasses MoCo-v2 by 5.1% on ImageNet-1K in linear evaluation and outperforms several other methods on transfer learning and object detection tasks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_06813 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | LeOCLR: Leveraging Original Images for Contrastive Learning of Visual Representations Alkhalefi, Mohammad Leontidis, Georgios Zhong, Mingjun Computer Vision and Pattern Recognition Contrastive instance discrimination methods outperform supervised learning in downstream tasks such as image classification and object detection. However, these methods rely heavily on data augmentation during representation learning, which can lead to suboptimal results if not implemented carefully. A common augmentation technique in contrastive learning is random cropping followed by resizing. This can degrade the quality of representation learning when the two random crops contain distinct semantic content. To tackle this issue, we introduce LeOCLR (Leveraging Original Images for Contrastive Learning of Visual Representations), a framework that employs a novel instance discrimination approach and an adapted loss function. This method prevents the loss of important semantic features caused by mapping different object parts during representation learning. Our experiments demonstrate that LeOCLR consistently improves representation learning across various datasets, outperforming baseline models. For instance, LeOCLR surpasses MoCo-v2 by 5.1% on ImageNet-1K in linear evaluation and outperforms several other methods on transfer learning and object detection tasks. |
| title | LeOCLR: Leveraging Original Images for Contrastive Learning of Visual Representations |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2403.06813 |