ETL: Efficient Transfer Learning for Face Tasks

Thrupthi Ann John, Isha Dua, Vineeth N. Balasubramanian, C. V. Jawahar

2022

Abstract

Transfer learning is a popular method for obtaining deep trained models for data-scarce face tasks such as head pose and emotion. However, current transfer learning methods are inefficient and time-consuming as they do not fully account for the relationships between related tasks. Moreover, the transferred model is large and computationally expensive. As an alternative, we propose ETL: a technique that efficiently transfers a pre-trained model to a new task by retaining only cross-task aware filters, resulting in a sparse transferred model. We demonstrate the effectiveness of ETL by transferring VGGFace, a popular face recognition model to four diverse face tasks. Our experiments show that we attain a size reduction up to 97% and an inference time reduction up to 94% while retaining 99.5% of the baseline transfer learning accuracy.

Paper Citation

in Harvard Style

John T., Dua I., Balasubramanian V. and Jawahar C. (2022). ETL: Efficient Transfer Learning for Face Tasks. In Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2022) - Volume 5: VISAPP; ISBN 978-989-758-555-5, SciTePress, pages 248-257. DOI: 10.5220/0010907700003124

in Bibtex Style

@conference{visapp22,
author={Thrupthi Ann John and Isha Dua and Vineeth N. Balasubramanian and C. V. Jawahar},
title={ETL: Efficient Transfer Learning for Face Tasks},
booktitle={Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2022) - Volume 5: VISAPP},
year={2022},
pages={248-257},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0010907700003124},
isbn={978-989-758-555-5},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2022) - Volume 5: VISAPP
TI - ETL: Efficient Transfer Learning for Face Tasks
SN - 978-989-758-555-5
AU - John T.
AU - Dua I.
AU - Balasubramanian V.
AU - Jawahar C.
PY - 2022
SP - 248
EP - 257
DO - 10.5220/0010907700003124
PB - SciTePress