Feature-based Lucas-Kanade and Active Appearance Models

Lucas-Kanade and Active Appearance Models are among the most commonly used methods for image alignment and facial fitting, respectively. They both utilize non-linear gradient descent, which is usually applied on intensity values. In this paper, we propose the employment of highly-descriptive, densel...

Full description

Bibliographic Details
Main Authors: Antonakos, Epameinondas, Alabort-i-Medina, Joan, Tzimiropoulos, Georgios, Zafeiriou, Stefanos P.
Format: Article
Published: Institute of Electrical and Electronics Engineers 2015
Online Access:https://eprints.nottingham.ac.uk/31444/
_version_ 1848794204199714816
author Antonakos, Epameinondas
Alabort-i-Medina, Joan
Tzimiropoulos, Georgios
Zafeiriou, Stefanos P.
author_facet Antonakos, Epameinondas
Alabort-i-Medina, Joan
Tzimiropoulos, Georgios
Zafeiriou, Stefanos P.
author_sort Antonakos, Epameinondas
building Nottingham Research Data Repository
collection Online Access
description Lucas-Kanade and Active Appearance Models are among the most commonly used methods for image alignment and facial fitting, respectively. They both utilize non-linear gradient descent, which is usually applied on intensity values. In this paper, we propose the employment of highly-descriptive, densely-sampled image features for both problems. We show that the strategy of warping the multi-channel dense feature image at each iteration is more beneficial than extracting features after warping the intensity image at each iteration. Motivated by this observation, we demonstrate robust and accurate alignment and fitting performance using a variety of powerful feature descriptors. Especially with the employment of HOG and SIFT features, our method significantly outperforms the current state-of-the-art results on in-the-wild databases.
first_indexed 2025-11-14T19:12:28Z
format Article
id nottingham-31444
institution University of Nottingham Malaysia Campus
institution_category Local University
last_indexed 2025-11-14T19:12:28Z
publishDate 2015
publisher Institute of Electrical and Electronics Engineers
recordtype eprints
repository_type Digital Repository
spelling nottingham-314442020-05-04T17:08:43Z https://eprints.nottingham.ac.uk/31444/ Feature-based Lucas-Kanade and Active Appearance Models Antonakos, Epameinondas Alabort-i-Medina, Joan Tzimiropoulos, Georgios Zafeiriou, Stefanos P. Lucas-Kanade and Active Appearance Models are among the most commonly used methods for image alignment and facial fitting, respectively. They both utilize non-linear gradient descent, which is usually applied on intensity values. In this paper, we propose the employment of highly-descriptive, densely-sampled image features for both problems. We show that the strategy of warping the multi-channel dense feature image at each iteration is more beneficial than extracting features after warping the intensity image at each iteration. Motivated by this observation, we demonstrate robust and accurate alignment and fitting performance using a variety of powerful feature descriptors. Especially with the employment of HOG and SIFT features, our method significantly outperforms the current state-of-the-art results on in-the-wild databases. Institute of Electrical and Electronics Engineers 2015-05-08 Article PeerReviewed Antonakos, Epameinondas, Alabort-i-Medina, Joan, Tzimiropoulos, Georgios and Zafeiriou, Stefanos P. (2015) Feature-based Lucas-Kanade and Active Appearance Models. IEEE Transactions on Image Processing, 24 (9). pp. 2617-2632. ISSN 1941-0042 http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=7104116 doi:10.1109/TIP.2015.2431445 doi:10.1109/TIP.2015.2431445
spellingShingle Antonakos, Epameinondas
Alabort-i-Medina, Joan
Tzimiropoulos, Georgios
Zafeiriou, Stefanos P.
Feature-based Lucas-Kanade and Active Appearance Models
title Feature-based Lucas-Kanade and Active Appearance Models
title_full Feature-based Lucas-Kanade and Active Appearance Models
title_fullStr Feature-based Lucas-Kanade and Active Appearance Models
title_full_unstemmed Feature-based Lucas-Kanade and Active Appearance Models
title_short Feature-based Lucas-Kanade and Active Appearance Models
title_sort feature-based lucas-kanade and active appearance models
url https://eprints.nottingham.ac.uk/31444/
https://eprints.nottingham.ac.uk/31444/
https://eprints.nottingham.ac.uk/31444/