arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2304.05554cs.CVcs.AI
收藏

从多模态信息监督中学习可迁移的行人表征

Learning Transferable Pedestrian Representation from Multimodal Information Supervision

University of Science and Technology of China(中国科学技术大学);Huawei Cloud(华为云)

Liping Bao, Longhui Wei, Xiaoyu Qiu, Wengang Zhou, Houqiang Li, Qi Tian

AI总结:

本文提出VAL-PAT框架,通过自监督对比学习、图像-文本对比学习和多属性分类三种目标,在LUPerson-TA数据集上预训练学习可迁移的行人表征,并有效迁移到行人重识别、属性识别和文本行人搜索等多个下游任务。

摘要:

近期关于无监督行人重识别(reID)的研究表明,在未标注的行人图像上进行预训练,相较于在ImageNet上预训练,能在下游reID任务上取得更优的性能。然而,这些预训练方法专为reID设计,难以灵活适配其他行人分析任务。本文提出VAL-PAT,一个利用多模态信息学习可迁移表征以增强各类行人分析任务的新框架。为训练该框架,我们引入三个学习目标,即自监督对比学习、图像-文本对比学习和多属性分类。自监督对比学习有助于学习行人的内在属性,图像-文本对比学习引导模型关注行人的外观信息,同时多属性分类促使模型识别属性以挖掘细粒度行人信息。我们首先在LUPerson-TA数据集上进行预训练,该数据集中每张图像均包含文本和属性标注,然后将学习到的表征迁移到多个下游任务,包括行人重识别、行人属性识别和基于文本的行人搜索。大量实验表明,该框架有助于学习通用的行人表征,从而在各类行人分析任务上取得有前景的结果。

英文摘要:

Recent researches on unsupervised person re-identification~(reID) have demonstrated that pre-training on unlabeled person images achieves superior performance on downstream reID tasks than pre-training on ImageNet. However, those pre-trained methods are specifically designed for reID and suffer flexible adaption to other pedestrian analysis tasks. In this paper, we propose VAL-PAT, a novel framework that learns transferable representations to enhance various pedestrian analysis tasks with multimodal information. To train our framework, we introduce three learning objectives, \emph{i.e.,} self-supervised contrastive learning, image-text contrastive learning and multi-attribute classification. The self-supervised contrastive learning facilitates the learning of the intrinsic pedestrian properties, while the image-text contrastive learning guides the model to focus on the appearance information of pedestrians.Meanwhile, multi-attribute classification encourages the model to recognize attributes to excavate fine-grained pedestrian information. We first perform pre-training on LUPerson-TA dataset, where each image contains text and attribute annotations, and then transfer the learned representations to various downstream tasks, including person reID, person attribute recognition and text-based person search. Extensive experiments demonstrate that our framework facilitates the learning of general pedestrian representations and thus leads to promising results on various pedestrian analysis tasks.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑