Adaptive Part Mining for Robust Visual Tracking

Visual tracking aims to estimate object state in a video sequence, which is challenging when facing drastic appearance changes. Most existing trackers conduct tracking with divided parts to handle appearance variations. However, these trackers commonly divide target objects into regular patches by a...

Ausführliche Beschreibung

Bibliographische Detailangaben
Veröffentlicht in:IEEE transactions on pattern analysis and machine intelligence. - 1979. - 45(2023), 10 vom: 16. Okt., Seite 11443-11457
1. Verfasser: Ma, Yinchao (VerfasserIn)
Weitere Verfasser: He, Jianfeng, Yang, Dawei, Zhang, Tianzhu, Wu, Feng
Format: Online-Aufsatz
Sprache:English
Veröffentlicht: 2023
Zugriff auf das übergeordnete Werk:IEEE transactions on pattern analysis and machine intelligence
Schlagworte:Journal Article
LEADER 01000naa a22002652 4500
001 NLM356946606
003 DE-627
005 20231226071532.0
007 cr uuu---uuuuu
008 231226s2023 xx |||||o 00| ||eng c
024 7 |a 10.1109/TPAMI.2023.3275034  |2 doi 
028 5 2 |a pubmed24n1189.xml 
035 |a (DE-627)NLM356946606 
035 |a (NLM)37192025 
040 |a DE-627  |b ger  |c DE-627  |e rakwb 
041 |a eng 
100 1 |a Ma, Yinchao  |e verfasserin  |4 aut 
245 1 0 |a Adaptive Part Mining for Robust Visual Tracking 
264 1 |c 2023 
336 |a Text  |b txt  |2 rdacontent 
337 |a ƒaComputermedien  |b c  |2 rdamedia 
338 |a ƒa Online-Ressource  |b cr  |2 rdacarrier 
500 |a Date Revised 06.09.2023 
500 |a published: Print-Electronic 
500 |a Citation Status PubMed-not-MEDLINE 
520 |a Visual tracking aims to estimate object state in a video sequence, which is challenging when facing drastic appearance changes. Most existing trackers conduct tracking with divided parts to handle appearance variations. However, these trackers commonly divide target objects into regular patches by a hand-designed splitting way, which is too coarse to align object parts well. Besides, a fixed part detector is difficult to partition targets with arbitrary categories and deformations. To address the above issues, we propose a novel adaptive part mining tracker (APMT) for robust tracking via a transformer architecture, including an object representation encoder, an adaptive part mining decoder, and an object state estimation decoder. The proposed APMT enjoys several merits. First, in the object representation encoder, object representation is learned by distinguishing target object from background regions. Second, in the adaptive part mining decoder, we introduce multiple part prototypes to adaptively capture target parts through cross-attention mechanisms for arbitrary categories and deformations. Third, in the object state estimation decoder, we propose two novel strategies to effectively handle appearance variations and distractors. Extensive experimental results demonstrate that our APMT achieves promising results with high FPS. Notably, our tracker is ranked the first place in the VOT-STb2022 challenge 
650 4 |a Journal Article 
700 1 |a He, Jianfeng  |e verfasserin  |4 aut 
700 1 |a Yang, Dawei  |e verfasserin  |4 aut 
700 1 |a Zhang, Tianzhu  |e verfasserin  |4 aut 
700 1 |a Wu, Feng  |e verfasserin  |4 aut 
773 0 8 |i Enthalten in  |t IEEE transactions on pattern analysis and machine intelligence  |d 1979  |g 45(2023), 10 vom: 16. Okt., Seite 11443-11457  |w (DE-627)NLM098212257  |x 1939-3539  |7 nnns 
773 1 8 |g volume:45  |g year:2023  |g number:10  |g day:16  |g month:10  |g pages:11443-11457 
856 4 0 |u http://dx.doi.org/10.1109/TPAMI.2023.3275034  |3 Volltext 
912 |a GBV_USEFLAG_A 
912 |a SYSFLAG_A 
912 |a GBV_NLM 
912 |a GBV_ILN_350 
951 |a AR 
952 |d 45  |j 2023  |e 10  |b 16  |c 10  |h 11443-11457