Meta Öǧrenme ile Video Nesne Doǧrulama

Translated title of the contribution: Video Object Verification via Meta-learning

Irem Beyza Onur, Filiz Gurkan, Bilge Gunsel

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Citation (Scopus)

Abstract

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.

Translated title of the contributionVideo Object Verification via Meta-learning
Original languageTurkish
Title of host publication2022 30th Signal Processing and Communications Applications Conference, SIU 2022
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781665450928
DOIs
Publication statusPublished - 2022
Event30th Signal Processing and Communications Applications Conference, SIU 2022 - Safranbolu, Turkey
Duration: 15 May 202218 May 2022

Publication series

Name2022 30th Signal Processing and Communications Applications Conference, SIU 2022

Conference

Conference30th Signal Processing and Communications Applications Conference, SIU 2022
Country/TerritoryTurkey
CitySafranbolu
Period15/05/2218/05/22

Bibliographical note

Publisher Copyright:
© 2022 IEEE.

Fingerprint

Dive into the research topics of 'Video Object Verification via Meta-learning'. Together they form a unique fingerprint.

Cite this