Visual Tracking by Adaptive Continual Meta-Learningopen access
- Authors
- Choi, Janghoon; Baik, Sungyong; Choi, Myungsub; Kwon, Junseok; Lee, Kyoung Mu
- Issue Date
- Jan-2022
- Publisher
- IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
- Keywords
- Visualization; Target tracking; Adaptation models; Training; Knowledge engineering; Classification algorithms; Task analysis; Continual learning; meta learning; object tracking; visual tracking
- Citation
- IEEE ACCESS, v.10, pp.9022 - 9035
- Indexed
- SCIE
SCOPUS
- Journal Title
- IEEE ACCESS
- Volume
- 10
- Start Page
- 9022
- End Page
- 9035
- URI
- https://scholarworks.bwise.kr/hanyang/handle/2021.sw.hanyang/188640
- DOI
- 10.1109/ACCESS.2022.3143809
- ISSN
- 2169-3536
- Abstract
- We formulate the visual tracking problem as a semi-supervised continual learning problem, where only an initial frame is labeled. In contrast to conventional meta-learning based approaches that regard visual tracking as an instance detection problem with a focus on finding good weights for model initialization, we consider both initialization and online update processes simultaneously under our adaptive continual meta-learning framework. The proposed adaptive meta-learning strategy dynamically generates the hyperparameters needed for fast initialization and online update to achieve more robustness via adaptively regulating the learning process. In addition, our continual meta-learning approach based on knowledge distillation scheme helps the tracker adapt to new examples while retaining its knowledge on previously seen examples. We apply our proposed framework to deep learning-based tracking algorithm to obtain noticeable performance gains and competitive results against recent state-of-the-art tracking algorithms while performing at real-time speeds.
- Files in This Item
-
Go to Link
- Appears in
Collections - ETC > 1. Journal Articles
![qrcode](https://api.qrserver.com/v1/create-qr-code/?size=55x55&data=https://scholarworks.bwise.kr/hanyang/handle/2021.sw.hanyang/188640)
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.