Beyond frontal faces: Improving Person Recognition using multiple cues

Ning Zhang, Manohar Paluri, Yaniv Taigman, Robert Fergus, Lubomir Bourdev

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

Abstract

We explore the task of recognizing peoples' identities in photo albums in an unconstrained setting. To facilitate this, we introduce the new People In Photo Albums (PIPA) dataset, consisting of over 60000 instances of ∼2000 individuals collected from public Flickr photo albums. With only about half of the person images containing a frontal face, the recognition task is very challenging due to the large variations in pose, clothing, camera viewpoint, image resolution and illumination. We propose the Pose Invariant PErson Recognition (PIPER) method, which accumulates the cues of poselet-level person recognizers trained by deep convolutional networks to discount for the pose variations, combined with a face recognizer and a global recognizer. Experiments on three different settings confirm that in our unconstrained setup PIPER significantly improves on the performance of DeepFace, which is one of the best face recognizers as measured on the LFW dataset.

Original languageEnglish (US)
Title of host publicationIEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015
PublisherIEEE Computer Society
Pages4804-4813
Number of pages10
Volume07-12-June-2015
ISBN (Print)9781467369640
DOIs
StatePublished - Oct 14 2015
EventIEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015 - Boston, United States
Duration: Jun 7 2015Jun 12 2015

Other

OtherIEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015
CountryUnited States
CityBoston
Period6/7/156/12/15

Fingerprint

Image resolution
Lighting
Cameras
Experiments

ASJC Scopus subject areas

  • Software
  • Computer Vision and Pattern Recognition

Cite this

Zhang, N., Paluri, M., Taigman, Y., Fergus, R., & Bourdev, L. (2015). Beyond frontal faces: Improving Person Recognition using multiple cues. In IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015 (Vol. 07-12-June-2015, pp. 4804-4813). [7299113] IEEE Computer Society. https://doi.org/10.1109/CVPR.2015.7299113

Beyond frontal faces : Improving Person Recognition using multiple cues. / Zhang, Ning; Paluri, Manohar; Taigman, Yaniv; Fergus, Robert; Bourdev, Lubomir.

IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015. Vol. 07-12-June-2015 IEEE Computer Society, 2015. p. 4804-4813 7299113.

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

Zhang, N, Paluri, M, Taigman, Y, Fergus, R & Bourdev, L 2015, Beyond frontal faces: Improving Person Recognition using multiple cues. in IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015. vol. 07-12-June-2015, 7299113, IEEE Computer Society, pp. 4804-4813, IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015, Boston, United States, 6/7/15. https://doi.org/10.1109/CVPR.2015.7299113
Zhang N, Paluri M, Taigman Y, Fergus R, Bourdev L. Beyond frontal faces: Improving Person Recognition using multiple cues. In IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015. Vol. 07-12-June-2015. IEEE Computer Society. 2015. p. 4804-4813. 7299113 https://doi.org/10.1109/CVPR.2015.7299113
Zhang, Ning ; Paluri, Manohar ; Taigman, Yaniv ; Fergus, Robert ; Bourdev, Lubomir. / Beyond frontal faces : Improving Person Recognition using multiple cues. IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015. Vol. 07-12-June-2015 IEEE Computer Society, 2015. pp. 4804-4813
@inproceedings{86b2fb70b2ba42adacf4b2ac7ea1b1b2,
title = "Beyond frontal faces: Improving Person Recognition using multiple cues",
abstract = "We explore the task of recognizing peoples' identities in photo albums in an unconstrained setting. To facilitate this, we introduce the new People In Photo Albums (PIPA) dataset, consisting of over 60000 instances of ∼2000 individuals collected from public Flickr photo albums. With only about half of the person images containing a frontal face, the recognition task is very challenging due to the large variations in pose, clothing, camera viewpoint, image resolution and illumination. We propose the Pose Invariant PErson Recognition (PIPER) method, which accumulates the cues of poselet-level person recognizers trained by deep convolutional networks to discount for the pose variations, combined with a face recognizer and a global recognizer. Experiments on three different settings confirm that in our unconstrained setup PIPER significantly improves on the performance of DeepFace, which is one of the best face recognizers as measured on the LFW dataset.",
author = "Ning Zhang and Manohar Paluri and Yaniv Taigman and Robert Fergus and Lubomir Bourdev",
year = "2015",
month = "10",
day = "14",
doi = "10.1109/CVPR.2015.7299113",
language = "English (US)",
isbn = "9781467369640",
volume = "07-12-June-2015",
pages = "4804--4813",
booktitle = "IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015",
publisher = "IEEE Computer Society",

}

TY - GEN

T1 - Beyond frontal faces

T2 - Improving Person Recognition using multiple cues

AU - Zhang, Ning

AU - Paluri, Manohar

AU - Taigman, Yaniv

AU - Fergus, Robert

AU - Bourdev, Lubomir

PY - 2015/10/14

Y1 - 2015/10/14

N2 - We explore the task of recognizing peoples' identities in photo albums in an unconstrained setting. To facilitate this, we introduce the new People In Photo Albums (PIPA) dataset, consisting of over 60000 instances of ∼2000 individuals collected from public Flickr photo albums. With only about half of the person images containing a frontal face, the recognition task is very challenging due to the large variations in pose, clothing, camera viewpoint, image resolution and illumination. We propose the Pose Invariant PErson Recognition (PIPER) method, which accumulates the cues of poselet-level person recognizers trained by deep convolutional networks to discount for the pose variations, combined with a face recognizer and a global recognizer. Experiments on three different settings confirm that in our unconstrained setup PIPER significantly improves on the performance of DeepFace, which is one of the best face recognizers as measured on the LFW dataset.

AB - We explore the task of recognizing peoples' identities in photo albums in an unconstrained setting. To facilitate this, we introduce the new People In Photo Albums (PIPA) dataset, consisting of over 60000 instances of ∼2000 individuals collected from public Flickr photo albums. With only about half of the person images containing a frontal face, the recognition task is very challenging due to the large variations in pose, clothing, camera viewpoint, image resolution and illumination. We propose the Pose Invariant PErson Recognition (PIPER) method, which accumulates the cues of poselet-level person recognizers trained by deep convolutional networks to discount for the pose variations, combined with a face recognizer and a global recognizer. Experiments on three different settings confirm that in our unconstrained setup PIPER significantly improves on the performance of DeepFace, which is one of the best face recognizers as measured on the LFW dataset.

UR - http://www.scopus.com/inward/record.url?scp=84959223360&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84959223360&partnerID=8YFLogxK

U2 - 10.1109/CVPR.2015.7299113

DO - 10.1109/CVPR.2015.7299113

M3 - Conference contribution

AN - SCOPUS:84959223360

SN - 9781467369640

VL - 07-12-June-2015

SP - 4804

EP - 4813

BT - IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015

PB - IEEE Computer Society

ER -