Boosting-based transfer learning for multi-view head-pose classification from surveillance videos

Radu L. Vieriu; Anoop K. Rajagopal; Ramanathan Subramanian; Oswald Lanz; Elisa Ricci; Nicu Sebe; Kalpathi Ramakrishnan

Boosting-based transfer learning for multi-view head-pose classification from surveillance videos

Radu L. Vieriu, Anoop K. Rajagopal, Ramanathan Subramanian, Oswald Lanz, Elisa Ricci, Nicu Sebe, Kalpathi Ramakrishnan

Research output: A Conference proceeding or a Chapter in Book › Conference contribution › peer-review

Abstract

This work proposes a boosting-based transfer learning approach for head-pose classification from multiple, low-resolution views. Head-pose classification performance is adversely affected when the source (training) and target (test) data arise from different distributions (due to change in face appearance, lighting, etc). Under such conditions, we employ Xferboost, a Logitboost-based transfer learning framework that integrates knowledge from a few labeled target samples with the source model to effectively minimize misclassifications on the target data. Experiments confirm that the Xferboost framework can improve classification performance by up to 6%, when knowledge is transferred between the CLEAR and FBK four-view headpose datasets.

Original language	English
Title of host publication	Proceedings of the 20th European Signal Processing Conference, EUSIPCO 2012
Editors	Béatrice Pesquet-Popescu, Corneliu Burileanu
Place of Publication	United States
Publisher	IEEE, Institute of Electrical and Electronics Engineers
Pages	649-653
Number of pages	5
ISBN (Print)	9781467310680
Publication status	Published - 2012
Externally published	Yes
Event	20th European Signal Processing Conference, EUSIPCO 2012 - Bucharest, Romania Duration: 27 Aug 2012 → 31 Aug 2012

Publication series

Name	European Signal Processing Conference
ISSN (Print)	2219-5491

Conference

Conference	20th European Signal Processing Conference, EUSIPCO 2012
Country/Territory	Romania
City	Bucharest
Period	27/08/12 → 31/08/12

Cite this

Vieriu, R. L., Rajagopal, A. K., Subramanian, R., Lanz, O., Ricci, E., Sebe, N., & Ramakrishnan, K. (2012). Boosting-based transfer learning for multi-view head-pose classification from surveillance videos. In B. Pesquet-Popescu, & C. Burileanu (Eds.), Proceedings of the 20th European Signal Processing Conference, EUSIPCO 2012 (pp. 649-653). [6333884] (European Signal Processing Conference). IEEE, Institute of Electrical and Electronics Engineers.

Vieriu, Radu L. ; Rajagopal, Anoop K. ; Subramanian, Ramanathan et al. / Boosting-based transfer learning for multi-view head-pose classification from surveillance videos. Proceedings of the 20th European Signal Processing Conference, EUSIPCO 2012. editor / Béatrice Pesquet-Popescu ; Corneliu Burileanu. United States : IEEE, Institute of Electrical and Electronics Engineers, 2012. pp. 649-653 (European Signal Processing Conference).

@inproceedings{b450c3f623ec4bcfaa6bc86a3693e0eb,

title = "Boosting-based transfer learning for multi-view head-pose classification from surveillance videos",

abstract = "This work proposes a boosting-based transfer learning approach for head-pose classification from multiple, low-resolution views. Head-pose classification performance is adversely affected when the source (training) and target (test) data arise from different distributions (due to change in face appearance, lighting, etc). Under such conditions, we employ Xferboost, a Logitboost-based transfer learning framework that integrates knowledge from a few labeled target samples with the source model to effectively minimize misclassifications on the target data. Experiments confirm that the Xferboost framework can improve classification performance by up to 6%, when knowledge is transferred between the CLEAR and FBK four-view headpose datasets.",

keywords = "boosting-based transfer learning, low-resolution, Multi-view headpose classification, Xferboost",

author = "Vieriu, {Radu L.} and Rajagopal, {Anoop K.} and Ramanathan Subramanian and Oswald Lanz and Elisa Ricci and Nicu Sebe and Kalpathi Ramakrishnan",

year = "2012",

language = "English",

isbn = "9781467310680",

series = "European Signal Processing Conference",

publisher = "IEEE, Institute of Electrical and Electronics Engineers",

pages = "649--653",

editor = "B{\'e}atrice Pesquet-Popescu and Corneliu Burileanu",

booktitle = "Proceedings of the 20th European Signal Processing Conference, EUSIPCO 2012",

address = "United States",

note = "20th European Signal Processing Conference, EUSIPCO 2012 ; Conference date: 27-08-2012 Through 31-08-2012",

}

Vieriu, RL, Rajagopal, AK, Subramanian, R, Lanz, O, Ricci, E, Sebe, N & Ramakrishnan, K 2012, Boosting-based transfer learning for multi-view head-pose classification from surveillance videos. in B Pesquet-Popescu & C Burileanu (eds), Proceedings of the 20th European Signal Processing Conference, EUSIPCO 2012., 6333884, European Signal Processing Conference, IEEE, Institute of Electrical and Electronics Engineers, United States, pp. 649-653, 20th European Signal Processing Conference, EUSIPCO 2012, Bucharest, Romania, 27/08/12.

Boosting-based transfer learning for multi-view head-pose classification from surveillance videos. / Vieriu, Radu L.; Rajagopal, Anoop K.; Subramanian, Ramanathan et al.
Proceedings of the 20th European Signal Processing Conference, EUSIPCO 2012. ed. / Béatrice Pesquet-Popescu; Corneliu Burileanu. United States: IEEE, Institute of Electrical and Electronics Engineers, 2012. p. 649-653 6333884 (European Signal Processing Conference).

Research output: A Conference proceeding or a Chapter in Book › Conference contribution › peer-review

TY - GEN

T1 - Boosting-based transfer learning for multi-view head-pose classification from surveillance videos

AU - Vieriu, Radu L.

AU - Rajagopal, Anoop K.

AU - Subramanian, Ramanathan

AU - Lanz, Oswald

AU - Ricci, Elisa

AU - Sebe, Nicu

AU - Ramakrishnan, Kalpathi

PY - 2012

Y1 - 2012

N2 - This work proposes a boosting-based transfer learning approach for head-pose classification from multiple, low-resolution views. Head-pose classification performance is adversely affected when the source (training) and target (test) data arise from different distributions (due to change in face appearance, lighting, etc). Under such conditions, we employ Xferboost, a Logitboost-based transfer learning framework that integrates knowledge from a few labeled target samples with the source model to effectively minimize misclassifications on the target data. Experiments confirm that the Xferboost framework can improve classification performance by up to 6%, when knowledge is transferred between the CLEAR and FBK four-view headpose datasets.

AB - This work proposes a boosting-based transfer learning approach for head-pose classification from multiple, low-resolution views. Head-pose classification performance is adversely affected when the source (training) and target (test) data arise from different distributions (due to change in face appearance, lighting, etc). Under such conditions, we employ Xferboost, a Logitboost-based transfer learning framework that integrates knowledge from a few labeled target samples with the source model to effectively minimize misclassifications on the target data. Experiments confirm that the Xferboost framework can improve classification performance by up to 6%, when knowledge is transferred between the CLEAR and FBK four-view headpose datasets.

KW - boosting-based transfer learning

KW - low-resolution

KW - Multi-view headpose classification

KW - Xferboost

UR - http://www.scopus.com/inward/record.url?scp=84869751759&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84869751759

SN - 9781467310680

T3 - European Signal Processing Conference

SP - 649

EP - 653

BT - Proceedings of the 20th European Signal Processing Conference, EUSIPCO 2012

A2 - Pesquet-Popescu, Béatrice

A2 - Burileanu, Corneliu

PB - IEEE, Institute of Electrical and Electronics Engineers

CY - United States

T2 - 20th European Signal Processing Conference, EUSIPCO 2012

Y2 - 27 August 2012 through 31 August 2012

ER -

Vieriu RL, Rajagopal AK, Subramanian R, Lanz O, Ricci E, Sebe N et al. Boosting-based transfer learning for multi-view head-pose classification from surveillance videos. In Pesquet-Popescu B, Burileanu C, editors, Proceedings of the 20th European Signal Processing Conference, EUSIPCO 2012. United States: IEEE, Institute of Electrical and Electronics Engineers. 2012. p. 649-653. 6333884. (European Signal Processing Conference).

Boosting-based transfer learning for multi-view head-pose classification from surveillance videos

Abstract

Publication series

Conference

Other files and links

Fingerprint

Cite this