Cascaded continuous regression for real-time incremental face tracking

This paper introduces a novel real-time algorithm for facial landmark tracking. Compared to detection, tracking has both additional challenges and opportunities. Arguably the most important aspect in this domain is updating a tracker's models as tracking progresses, also known as incremental (f...

Full description

Bibliographic Details
Main Authors: Sánchez Lozano, Enrique, Martinez, Brais, Tzimiropoulos, Georgios, Valstar, Michel F.
Format: Conference or Workshop Item
Published: 2016
Online Access:https://eprints.nottingham.ac.uk/35721/
_version_ 1848795146492051456
author Sánchez Lozano, Enrique
Martinez, Brais
Tzimiropoulos, Georgios
Valstar, Michel F.
author_facet Sánchez Lozano, Enrique
Martinez, Brais
Tzimiropoulos, Georgios
Valstar, Michel F.
author_sort Sánchez Lozano, Enrique
building Nottingham Research Data Repository
collection Online Access
description This paper introduces a novel real-time algorithm for facial landmark tracking. Compared to detection, tracking has both additional challenges and opportunities. Arguably the most important aspect in this domain is updating a tracker's models as tracking progresses, also known as incremental (face) tracking. While this should result in more accurate localisation, how to do this online and in real time without causing a tracker to drift is still an important open research question. We address this question in the cascaded regression framework, the state-of-the-art approach for facial landmark localisation. Because incremental learning for cascaded regression is costly, we propose a much more efficient yet equally accurate alternative using continuous regression. More specifically, we first propose cascaded continuous regression (CCR) and show its accuracy is equivalent to the Supervised Descent Method. We then derive the incremental learning updates for CCR (iCCR) and show that it is an order of magnitude faster than standard incremental learning for cascaded regression, bringing the time required for the update from seconds down to a fraction of a second, thus enabling real-time tracking. Finally, we evaluate iCCR and show the importance of incremental learning in achieving state-of-the-art performance. Code for our iCCR is available from http://www.cs.nott.ac.uk/~psxes1.
first_indexed 2025-11-14T19:27:27Z
format Conference or Workshop Item
id nottingham-35721
institution University of Nottingham Malaysia Campus
institution_category Local University
last_indexed 2025-11-14T19:27:27Z
publishDate 2016
recordtype eprints
repository_type Digital Repository
spelling nottingham-357212020-05-04T18:17:16Z https://eprints.nottingham.ac.uk/35721/ Cascaded continuous regression for real-time incremental face tracking Sánchez Lozano, Enrique Martinez, Brais Tzimiropoulos, Georgios Valstar, Michel F. This paper introduces a novel real-time algorithm for facial landmark tracking. Compared to detection, tracking has both additional challenges and opportunities. Arguably the most important aspect in this domain is updating a tracker's models as tracking progresses, also known as incremental (face) tracking. While this should result in more accurate localisation, how to do this online and in real time without causing a tracker to drift is still an important open research question. We address this question in the cascaded regression framework, the state-of-the-art approach for facial landmark localisation. Because incremental learning for cascaded regression is costly, we propose a much more efficient yet equally accurate alternative using continuous regression. More specifically, we first propose cascaded continuous regression (CCR) and show its accuracy is equivalent to the Supervised Descent Method. We then derive the incremental learning updates for CCR (iCCR) and show that it is an order of magnitude faster than standard incremental learning for cascaded regression, bringing the time required for the update from seconds down to a fraction of a second, thus enabling real-time tracking. Finally, we evaluate iCCR and show the importance of incremental learning in achieving state-of-the-art performance. Code for our iCCR is available from http://www.cs.nott.ac.uk/~psxes1. 2016-10-13 Conference or Workshop Item PeerReviewed Sánchez Lozano, Enrique, Martinez, Brais, Tzimiropoulos, Georgios and Valstar, Michel F. (2016) Cascaded continuous regression for real-time incremental face tracking. In: 14th European Conference on Computer Vision (EECV 2016), 8-16 October 2016, Amsterdam, Netherlands. http://link.springer.com/chapter/10.1007/978-3-319-46484-8_39
spellingShingle Sánchez Lozano, Enrique
Martinez, Brais
Tzimiropoulos, Georgios
Valstar, Michel F.
Cascaded continuous regression for real-time incremental face tracking
title Cascaded continuous regression for real-time incremental face tracking
title_full Cascaded continuous regression for real-time incremental face tracking
title_fullStr Cascaded continuous regression for real-time incremental face tracking
title_full_unstemmed Cascaded continuous regression for real-time incremental face tracking
title_short Cascaded continuous regression for real-time incremental face tracking
title_sort cascaded continuous regression for real-time incremental face tracking
url https://eprints.nottingham.ac.uk/35721/
https://eprints.nottingham.ac.uk/35721/