Matches in SemOpenAlex for { <https://semopenalex.org/work/W4312532537> ?p ?o ?g. }
- W4312532537 endingPage "214" @default.
- W4312532537 startingPage "198" @default.
- W4312532537 abstract "AbstractIn person re-identification (ReID), very recent researches have validated pre-training the models on unlabelled person images is much better than on ImageNet. However, these researches directly apply the existing self-supervised learning (SSL) methods designed for image classification to ReID without any adaption in the framework. These SSL methods match the outputs of local views (e.g., red T-shirt, blue shorts) to those of the global views at the same time, losing lots of details. In this paper, we propose a ReID-specific pre-training method, Part-Aware Self-Supervised pre-training (PASS), which can generate part-level features to offer fine-grained information and is more suitable for ReID. PASS divides the images into several local areas, and the local views randomly cropped from each area are assigned a specific learnable [PART] token. On the other hand, the [PART]s of all local areas are also appended to the global views. PASS learns to match the outputs of the local views and global views on the same [PART]. That is, the learned [PART] of the local views from a local area is only matched with the corresponding [PART] learned from the global views. As a result, each [PART] can focus on a specific local area of the image and extracts fine-grained information of this area. Experiments show PASS sets the new state-of-the-art performances on Market1501 and MSMT17 on various ReID tasks, e.g., vanilla ViT-S/16 pre-trained by PASS achieves 92.2%/90.2%/88.5% mAP accuracy on Market1501 for supervised/UDA/USL ReID. Our codes are available at https://github.com/CASIA-IVA-Lab/PASS-reID.KeywordsPerson re-identificationSelf-supervised pre-trainingLocal representations" @default.
- W4312532537 created "2023-01-05" @default.
- W4312532537 creator A5000205902 @default.
- W4312532537 creator A5025994300 @default.
- W4312532537 creator A5027279261 @default.
- W4312532537 creator A5058420913 @default.
- W4312532537 creator A5065043476 @default.
- W4312532537 creator A5085707125 @default.
- W4312532537 date "2022-01-01" @default.
- W4312532537 modified "2023-10-14" @default.
- W4312532537 title "PASS: Part-Aware Self-Supervised Pre-Training for Person Re-Identification" @default.
- W4312532537 cites W2108598243 @default.
- W4312532537 cites W2194775991 @default.
- W4312532537 cites W2204750386 @default.
- W4312532537 cites W2884366600 @default.
- W4312532537 cites W2963047834 @default.
- W4312532537 cites W2963805953 @default.
- W4312532537 cites W2963842104 @default.
- W4312532537 cites W2967515867 @default.
- W4312532537 cites W2971098960 @default.
- W4312532537 cites W2990827756 @default.
- W4312532537 cites W3034607353 @default.
- W4312532537 cites W3035186652 @default.
- W4312532537 cites W3035402405 @default.
- W4312532537 cites W3035524453 @default.
- W4312532537 cites W3100506510 @default.
- W4312532537 cites W3108381568 @default.
- W4312532537 cites W3115484111 @default.
- W4312532537 cites W3145450063 @default.
- W4312532537 cites W3159481202 @default.
- W4312532537 cites W3175445198 @default.
- W4312532537 cites W3184225461 @default.
- W4312532537 cites W4214736485 @default.
- W4312532537 cites W4220906215 @default.
- W4312532537 cites W4226437068 @default.
- W4312532537 doi "https://doi.org/10.1007/978-3-031-19781-9_12" @default.
- W4312532537 hasPublicationYear "2022" @default.
- W4312532537 type Work @default.
- W4312532537 citedByCount "4" @default.
- W4312532537 countsByYear W43125325372023 @default.
- W4312532537 crossrefType "book-chapter" @default.
- W4312532537 hasAuthorship W4312532537A5000205902 @default.
- W4312532537 hasAuthorship W4312532537A5025994300 @default.
- W4312532537 hasAuthorship W4312532537A5027279261 @default.
- W4312532537 hasAuthorship W4312532537A5058420913 @default.
- W4312532537 hasAuthorship W4312532537A5065043476 @default.
- W4312532537 hasAuthorship W4312532537A5085707125 @default.
- W4312532537 hasBestOaLocation W43125325372 @default.
- W4312532537 hasConcept C115961682 @default.
- W4312532537 hasConcept C116834253 @default.
- W4312532537 hasConcept C119857082 @default.
- W4312532537 hasConcept C120665830 @default.
- W4312532537 hasConcept C121332964 @default.
- W4312532537 hasConcept C153180895 @default.
- W4312532537 hasConcept C153294291 @default.
- W4312532537 hasConcept C154945302 @default.
- W4312532537 hasConcept C192209626 @default.
- W4312532537 hasConcept C2777211547 @default.
- W4312532537 hasConcept C2994044699 @default.
- W4312532537 hasConcept C31972630 @default.
- W4312532537 hasConcept C33923547 @default.
- W4312532537 hasConcept C41008148 @default.
- W4312532537 hasConcept C59822182 @default.
- W4312532537 hasConcept C86803240 @default.
- W4312532537 hasConcept C94375191 @default.
- W4312532537 hasConceptScore W4312532537C115961682 @default.
- W4312532537 hasConceptScore W4312532537C116834253 @default.
- W4312532537 hasConceptScore W4312532537C119857082 @default.
- W4312532537 hasConceptScore W4312532537C120665830 @default.
- W4312532537 hasConceptScore W4312532537C121332964 @default.
- W4312532537 hasConceptScore W4312532537C153180895 @default.
- W4312532537 hasConceptScore W4312532537C153294291 @default.
- W4312532537 hasConceptScore W4312532537C154945302 @default.
- W4312532537 hasConceptScore W4312532537C192209626 @default.
- W4312532537 hasConceptScore W4312532537C2777211547 @default.
- W4312532537 hasConceptScore W4312532537C2994044699 @default.
- W4312532537 hasConceptScore W4312532537C31972630 @default.
- W4312532537 hasConceptScore W4312532537C33923547 @default.
- W4312532537 hasConceptScore W4312532537C41008148 @default.
- W4312532537 hasConceptScore W4312532537C59822182 @default.
- W4312532537 hasConceptScore W4312532537C86803240 @default.
- W4312532537 hasConceptScore W4312532537C94375191 @default.
- W4312532537 hasLocation W43125325371 @default.
- W4312532537 hasLocation W43125325372 @default.
- W4312532537 hasOpenAccess W4312532537 @default.
- W4312532537 hasPrimaryLocation W43125325371 @default.
- W4312532537 hasRelatedWork W1511650216 @default.
- W4312532537 hasRelatedWork W2004643608 @default.
- W4312532537 hasRelatedWork W2005185696 @default.
- W4312532537 hasRelatedWork W2141819204 @default.
- W4312532537 hasRelatedWork W2154757407 @default.
- W4312532537 hasRelatedWork W2161229648 @default.
- W4312532537 hasRelatedWork W2235753890 @default.
- W4312532537 hasRelatedWork W2891108385 @default.
- W4312532537 hasRelatedWork W2990667865 @default.
- W4312532537 hasRelatedWork W3007420330 @default.
- W4312532537 isParatext "false" @default.
- W4312532537 isRetracted "false" @default.