.png){kind=icon}
返回
DenseCL: A simple framework for self-supervised dense visual pre-training
.png){kind=icon}
OA
AI
DOI：10.1016/j.visinf.2022.09.003.png)
.png){kind=icon}
.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Self-supervised learning aims to learn a universal feature representation without labels. To date, most existing self-supervised learning methods are designed and optimized for image classification. These pre-trained models can be sub-optimal for dense prediction tasks due to the discrepancy between image-level prediction and pixel-level prediction. To fill this gap, we aim to design an effective, dense self-supervised learning framework that directly works at the level of pixels (or local features) by tak-ing into account the correspondence between local features. Specifically, we present dense contrastive learning (DenseCL), which implements self-supervised learning by optimizing a pairwise contrastive (dis)similarity loss at the pixel level between two views of input images. Compared to the supervised ImageNet pre-training and other self-supervised learning methods, our self-supervised DenseCL pre -training demonstrates consistently superior performance when transferring to downstream dense prediction tasks including object detection, semantic segmentation and instance segmentation. Specif-ically, our approach significantly outperforms the strong MoCo-v2 by 2.0% AP on PASCAL VOC object detection, 1.1% AP on COCO object detection, 0.9% AP on COCO instance segmentation, 3.0% mIoU on PASCAL VOC semantic segmentation and 1.8% mIoU on Cityscapes semantic segmentation. The improvements are up to 3.5% AP and 8.8% mIoU over MoCo-v2, and 6.1% AP and 6.1% mIoU over supervised counterpart with frozen-backbone evaluation protocol.Code and models are available at: https://git.io/DenseCL (c) 2023 The Authors. Published by Elsevier B.V. on behalf of Zhejiang University and Zhejiang University Press Co. Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Keyword:
Self-supervised learning
Visual pre-training
Dense prediction tasks
AI总结
对已上传原文的论文进行重点信息的提取，主要内容包括：简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。
期刊
IF:
3.9
论文数:
248
被引数:
628
机构
引用论文
Different Forms of Vigilance in Response to the Presence of Predators and Conspecifics in a Group‐Living Mammal, the European Rabbit.png)
Ethology
EthologyIF0
PREAI
分享收藏Perovskite solid electrolytes: Structure, transport properties and fuel cell applications钙钛矿固体电解质: 结构、输运性质和燃料电池应用
PREAI
分享收藏Thermal comfort, perceived air quality, and cognitive performance when personally controlled air movement is used by tropically acclimatized persons当热带适应的人使用个人控制的空气运动时，热舒适性，感知的空气质量和认知表现Indoor AirOAAI
Indoor AirIF0
OAAI分享收藏
Do Perceptions of Competence Mediate The Relationship Between Fundamental Motor Skill Proficiency and Physical Activity Levels of Children in Kindergarten?能力的感知是否可以介导幼儿园儿童的基本运动技能熟练程度与身体活动水平之间的关系？
PREAI
分享收藏
查看更多内容