Meta Öǧrenme ile Video Nesne Doǧrulama

Irem Beyza Onur, Filiz Gurkan, Bilge Gunsel

Araştırma sonucu: ???type-name???Konferans katkısıbilirkişi

1 Atıf (Scopus)

Özet

Performance of a long term object tracker relies on the object detection accuracy. Although several object detectors are proposed in the literature, robustness to target disappearances and reappearances is still a challenging problem. To deal with this problem, we propose an inference pipeline that integrates an object detector with a meta-learner, both locally trained. This is achieved by replacing the head classification layer of the object detector by a meta-learner that also enables verification of the target. In particular, Mask R-CNN object detector is integrated with SDNet trained end-to-end for object tracking. Improvement achieved by MAML++ meta learner trained as a classifier is also evaluated. Numerical results reported on VOT2020-LT long term video dataset demonstrate that both SDNet and MAML++ meta-learners improve the detection accuracy for unseen object classes. Moreover verification by SDNET provides 7% increase on detection of target disappearance and reappearance frames.

Tercüme edilen katkı başlığıVideo Object Verification via Meta-learning
Orijinal dilTürkçe
Ana bilgisayar yayını başlığı2022 30th Signal Processing and Communications Applications Conference, SIU 2022
YayınlayanInstitute of Electrical and Electronics Engineers Inc.
ISBN (Elektronik)9781665450928
DOI'lar
Yayın durumuYayınlandı - 2022
Etkinlik30th Signal Processing and Communications Applications Conference, SIU 2022 - Safranbolu, Turkey
Süre: 15 May 202218 May 2022

Yayın serisi

Adı2022 30th Signal Processing and Communications Applications Conference, SIU 2022

???event.eventtypes.event.conference???

???event.eventtypes.event.conference???30th Signal Processing and Communications Applications Conference, SIU 2022
Ülke/BölgeTurkey
ŞehirSafranbolu
Periyot15/05/2218/05/22

Bibliyografik not

Publisher Copyright:
© 2022 IEEE.

Keywords

  • meta-learning
  • object detection and verification

Parmak izi

Meta Öǧrenme ile Video Nesne Doǧrulama' araştırma başlıklarına git. Birlikte benzersiz bir parmak izi oluştururlar.

Alıntı Yap