Instance-based deep transfer learning

Tianyang Wang, Jun Huan, Michelle Zhu

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Deep transfer learning recently has acquired significant research interest. It makes use of pre-trained models that are learned from a source domain, and utilizes these models for the tasks in a target domain. Model-based deep transfer learning is probably the most frequently used method. However, very little research work has been devoted to enhancing deep transfer learning by focusing on the influence of data. In this paper, we propose an instance-based approach to improve deep transfer learning in a target domain. Specifically, we choose a pre-trained model from a source domain and apply this model to estimate the influence of training samples in a target domain. Then we optimize the training data of the target domain by removing the training samples that will lower the performance of the pre-trained model. We later either fine-tune the pre-trained model with the optimized training data in the target domain, or build a new model which is initialized partially based on the pre-trained model, and fine-tune it with the optimized training data in the target domain. Using this approach, transfer learning can help deep learning models to capture more useful features. Extensive experiments demonstrate the effectiveness of our approach on boosting the quality of deep learning models for some common computer vision tasks, such as image classification.

Original languageEnglish
Title of host publicationProceedings - 2019 IEEE Winter Conference on Applications of Computer Vision, WACV 2019
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages367-375
Number of pages9
ISBN (Electronic)9781728119755
DOIs
StatePublished - 4 Mar 2019
Event19th IEEE Winter Conference on Applications of Computer Vision, WACV 2019 - Waikoloa Village, United States
Duration: 7 Jan 201911 Jan 2019

Publication series

NameProceedings - 2019 IEEE Winter Conference on Applications of Computer Vision, WACV 2019

Conference

Conference19th IEEE Winter Conference on Applications of Computer Vision, WACV 2019
CountryUnited States
CityWaikoloa Village
Period7/01/1911/01/19

Fingerprint

Image classification
Computer vision
Experiments
Deep learning

Cite this

Wang, T., Huan, J., & Zhu, M. (2019). Instance-based deep transfer learning. In Proceedings - 2019 IEEE Winter Conference on Applications of Computer Vision, WACV 2019 (pp. 367-375). [8659197] (Proceedings - 2019 IEEE Winter Conference on Applications of Computer Vision, WACV 2019). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/WACV.2019.00045
Wang, Tianyang ; Huan, Jun ; Zhu, Michelle. / Instance-based deep transfer learning. Proceedings - 2019 IEEE Winter Conference on Applications of Computer Vision, WACV 2019. Institute of Electrical and Electronics Engineers Inc., 2019. pp. 367-375 (Proceedings - 2019 IEEE Winter Conference on Applications of Computer Vision, WACV 2019).
@inproceedings{5ebc4836f8374dbe84f326532bce5b83,
title = "Instance-based deep transfer learning",
abstract = "Deep transfer learning recently has acquired significant research interest. It makes use of pre-trained models that are learned from a source domain, and utilizes these models for the tasks in a target domain. Model-based deep transfer learning is probably the most frequently used method. However, very little research work has been devoted to enhancing deep transfer learning by focusing on the influence of data. In this paper, we propose an instance-based approach to improve deep transfer learning in a target domain. Specifically, we choose a pre-trained model from a source domain and apply this model to estimate the influence of training samples in a target domain. Then we optimize the training data of the target domain by removing the training samples that will lower the performance of the pre-trained model. We later either fine-tune the pre-trained model with the optimized training data in the target domain, or build a new model which is initialized partially based on the pre-trained model, and fine-tune it with the optimized training data in the target domain. Using this approach, transfer learning can help deep learning models to capture more useful features. Extensive experiments demonstrate the effectiveness of our approach on boosting the quality of deep learning models for some common computer vision tasks, such as image classification.",
author = "Tianyang Wang and Jun Huan and Michelle Zhu",
year = "2019",
month = "3",
day = "4",
doi = "10.1109/WACV.2019.00045",
language = "English",
series = "Proceedings - 2019 IEEE Winter Conference on Applications of Computer Vision, WACV 2019",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "367--375",
booktitle = "Proceedings - 2019 IEEE Winter Conference on Applications of Computer Vision, WACV 2019",

}

Wang, T, Huan, J & Zhu, M 2019, Instance-based deep transfer learning. in Proceedings - 2019 IEEE Winter Conference on Applications of Computer Vision, WACV 2019., 8659197, Proceedings - 2019 IEEE Winter Conference on Applications of Computer Vision, WACV 2019, Institute of Electrical and Electronics Engineers Inc., pp. 367-375, 19th IEEE Winter Conference on Applications of Computer Vision, WACV 2019, Waikoloa Village, United States, 7/01/19. https://doi.org/10.1109/WACV.2019.00045

Instance-based deep transfer learning. / Wang, Tianyang; Huan, Jun; Zhu, Michelle.

Proceedings - 2019 IEEE Winter Conference on Applications of Computer Vision, WACV 2019. Institute of Electrical and Electronics Engineers Inc., 2019. p. 367-375 8659197 (Proceedings - 2019 IEEE Winter Conference on Applications of Computer Vision, WACV 2019).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Instance-based deep transfer learning

AU - Wang, Tianyang

AU - Huan, Jun

AU - Zhu, Michelle

PY - 2019/3/4

Y1 - 2019/3/4

N2 - Deep transfer learning recently has acquired significant research interest. It makes use of pre-trained models that are learned from a source domain, and utilizes these models for the tasks in a target domain. Model-based deep transfer learning is probably the most frequently used method. However, very little research work has been devoted to enhancing deep transfer learning by focusing on the influence of data. In this paper, we propose an instance-based approach to improve deep transfer learning in a target domain. Specifically, we choose a pre-trained model from a source domain and apply this model to estimate the influence of training samples in a target domain. Then we optimize the training data of the target domain by removing the training samples that will lower the performance of the pre-trained model. We later either fine-tune the pre-trained model with the optimized training data in the target domain, or build a new model which is initialized partially based on the pre-trained model, and fine-tune it with the optimized training data in the target domain. Using this approach, transfer learning can help deep learning models to capture more useful features. Extensive experiments demonstrate the effectiveness of our approach on boosting the quality of deep learning models for some common computer vision tasks, such as image classification.

AB - Deep transfer learning recently has acquired significant research interest. It makes use of pre-trained models that are learned from a source domain, and utilizes these models for the tasks in a target domain. Model-based deep transfer learning is probably the most frequently used method. However, very little research work has been devoted to enhancing deep transfer learning by focusing on the influence of data. In this paper, we propose an instance-based approach to improve deep transfer learning in a target domain. Specifically, we choose a pre-trained model from a source domain and apply this model to estimate the influence of training samples in a target domain. Then we optimize the training data of the target domain by removing the training samples that will lower the performance of the pre-trained model. We later either fine-tune the pre-trained model with the optimized training data in the target domain, or build a new model which is initialized partially based on the pre-trained model, and fine-tune it with the optimized training data in the target domain. Using this approach, transfer learning can help deep learning models to capture more useful features. Extensive experiments demonstrate the effectiveness of our approach on boosting the quality of deep learning models for some common computer vision tasks, such as image classification.

UR - http://www.scopus.com/inward/record.url?scp=85063574777&partnerID=8YFLogxK

U2 - 10.1109/WACV.2019.00045

DO - 10.1109/WACV.2019.00045

M3 - Conference contribution

AN - SCOPUS:85063574777

T3 - Proceedings - 2019 IEEE Winter Conference on Applications of Computer Vision, WACV 2019

SP - 367

EP - 375

BT - Proceedings - 2019 IEEE Winter Conference on Applications of Computer Vision, WACV 2019

PB - Institute of Electrical and Electronics Engineers Inc.

ER -

Wang T, Huan J, Zhu M. Instance-based deep transfer learning. In Proceedings - 2019 IEEE Winter Conference on Applications of Computer Vision, WACV 2019. Institute of Electrical and Electronics Engineers Inc. 2019. p. 367-375. 8659197. (Proceedings - 2019 IEEE Winter Conference on Applications of Computer Vision, WACV 2019). https://doi.org/10.1109/WACV.2019.00045