Exploiting view-specific appearance similarities across classes for zero-shot pose prediction: A metric learning approach

Publikation: Beitrag in Buch/Bericht/Sammelwerk/KonferenzbandAufsatz in KonferenzbandForschungPeer-Review

Autoren

Externe Organisationen

  • Ulsan National Institute of Science and Technology
  • Carnegie Mellon University
Forschungs-netzwerk anzeigen

Details

OriginalspracheEnglisch
Titel des Sammelwerks30th AAAI Conference on Artificial Intelligence, AAAI 2016
Seiten3523-3529
Seitenumfang7
ISBN (elektronisch)9781577357605
PublikationsstatusVeröffentlicht - 2016
Veranstaltung30th AAAI Conference on Artificial Intelligence, AAAI 2016 - Phoenix, USA / Vereinigte Staaten
Dauer: 12 Feb. 201617 Feb. 2016

Publikationsreihe

Name30th AAAI Conference on Artificial Intelligence, AAAI 2016

Abstract

Viewpoint estimation, especially in case of multiple object classes, remains an important and challenging problem. First, objects under different views undergo extreme appearance variations, often making within-class variance larger than between-class variance. Second, obtaining precise ground truth for real-world images, necessary for training supervised viewpoint estimation models, is extremely difficult and time consuming. As a result, annotated data is often available only for a limited number of classes. Hence it is desirable to share viewpoint information across classes. Additional complexity arises from unaligned pose labels between classes, i.e. a side view of a car might look more like a frontal view of a toaster, than its side view. To address these problems, we propose a metric learning approach for joint class prediction and pose estimation. Our approach allows to circumvent the problem of viewpoint alignment across multiple classes, and does not require dense viewpoint labels. Moreover, we show, that the learned metric generalizes to new classes, for which the pose labels are not available, and therefore makes it possible to use only partially annotated training sets, relying on the intrinsic similarities in the viewpoint manifolds. We evaluate our approach on two challenging multi-class datasets, 3DObjects and PASCAL3D+.

ASJC Scopus Sachgebiete

Zitieren

Exploiting view-specific appearance similarities across classes for zero-shot pose prediction: A metric learning approach. / Kuznetsova, Alina; Hwang, Sung Ju; Rosenhahn, Bodo et al.
30th AAAI Conference on Artificial Intelligence, AAAI 2016. 2016. S. 3523-3529 (30th AAAI Conference on Artificial Intelligence, AAAI 2016).

Publikation: Beitrag in Buch/Bericht/Sammelwerk/KonferenzbandAufsatz in KonferenzbandForschungPeer-Review

Kuznetsova, A, Hwang, SJ, Rosenhahn, B & Sigal, L 2016, Exploiting view-specific appearance similarities across classes for zero-shot pose prediction: A metric learning approach. in 30th AAAI Conference on Artificial Intelligence, AAAI 2016. 30th AAAI Conference on Artificial Intelligence, AAAI 2016, S. 3523-3529, 30th AAAI Conference on Artificial Intelligence, AAAI 2016, Phoenix, USA / Vereinigte Staaten, 12 Feb. 2016.
Kuznetsova, A., Hwang, S. J., Rosenhahn, B., & Sigal, L. (2016). Exploiting view-specific appearance similarities across classes for zero-shot pose prediction: A metric learning approach. In 30th AAAI Conference on Artificial Intelligence, AAAI 2016 (S. 3523-3529). (30th AAAI Conference on Artificial Intelligence, AAAI 2016).
Kuznetsova A, Hwang SJ, Rosenhahn B, Sigal L. Exploiting view-specific appearance similarities across classes for zero-shot pose prediction: A metric learning approach. in 30th AAAI Conference on Artificial Intelligence, AAAI 2016. 2016. S. 3523-3529. (30th AAAI Conference on Artificial Intelligence, AAAI 2016).
Kuznetsova, Alina ; Hwang, Sung Ju ; Rosenhahn, Bodo et al. / Exploiting view-specific appearance similarities across classes for zero-shot pose prediction : A metric learning approach. 30th AAAI Conference on Artificial Intelligence, AAAI 2016. 2016. S. 3523-3529 (30th AAAI Conference on Artificial Intelligence, AAAI 2016).
Download
@inproceedings{3eb8838bfc32452dbcb930975d2ccf3a,
title = "Exploiting view-specific appearance similarities across classes for zero-shot pose prediction: A metric learning approach",
abstract = "Viewpoint estimation, especially in case of multiple object classes, remains an important and challenging problem. First, objects under different views undergo extreme appearance variations, often making within-class variance larger than between-class variance. Second, obtaining precise ground truth for real-world images, necessary for training supervised viewpoint estimation models, is extremely difficult and time consuming. As a result, annotated data is often available only for a limited number of classes. Hence it is desirable to share viewpoint information across classes. Additional complexity arises from unaligned pose labels between classes, i.e. a side view of a car might look more like a frontal view of a toaster, than its side view. To address these problems, we propose a metric learning approach for joint class prediction and pose estimation. Our approach allows to circumvent the problem of viewpoint alignment across multiple classes, and does not require dense viewpoint labels. Moreover, we show, that the learned metric generalizes to new classes, for which the pose labels are not available, and therefore makes it possible to use only partially annotated training sets, relying on the intrinsic similarities in the viewpoint manifolds. We evaluate our approach on two challenging multi-class datasets, 3DObjects and PASCAL3D+.",
author = "Alina Kuznetsova and Hwang, {Sung Ju} and Bodo Rosenhahn and Leonid Sigal",
year = "2016",
language = "English",
series = "30th AAAI Conference on Artificial Intelligence, AAAI 2016",
pages = "3523--3529",
booktitle = "30th AAAI Conference on Artificial Intelligence, AAAI 2016",
note = "30th AAAI Conference on Artificial Intelligence, AAAI 2016 ; Conference date: 12-02-2016 Through 17-02-2016",

}

Download

TY - GEN

T1 - Exploiting view-specific appearance similarities across classes for zero-shot pose prediction

T2 - 30th AAAI Conference on Artificial Intelligence, AAAI 2016

AU - Kuznetsova, Alina

AU - Hwang, Sung Ju

AU - Rosenhahn, Bodo

AU - Sigal, Leonid

PY - 2016

Y1 - 2016

N2 - Viewpoint estimation, especially in case of multiple object classes, remains an important and challenging problem. First, objects under different views undergo extreme appearance variations, often making within-class variance larger than between-class variance. Second, obtaining precise ground truth for real-world images, necessary for training supervised viewpoint estimation models, is extremely difficult and time consuming. As a result, annotated data is often available only for a limited number of classes. Hence it is desirable to share viewpoint information across classes. Additional complexity arises from unaligned pose labels between classes, i.e. a side view of a car might look more like a frontal view of a toaster, than its side view. To address these problems, we propose a metric learning approach for joint class prediction and pose estimation. Our approach allows to circumvent the problem of viewpoint alignment across multiple classes, and does not require dense viewpoint labels. Moreover, we show, that the learned metric generalizes to new classes, for which the pose labels are not available, and therefore makes it possible to use only partially annotated training sets, relying on the intrinsic similarities in the viewpoint manifolds. We evaluate our approach on two challenging multi-class datasets, 3DObjects and PASCAL3D+.

AB - Viewpoint estimation, especially in case of multiple object classes, remains an important and challenging problem. First, objects under different views undergo extreme appearance variations, often making within-class variance larger than between-class variance. Second, obtaining precise ground truth for real-world images, necessary for training supervised viewpoint estimation models, is extremely difficult and time consuming. As a result, annotated data is often available only for a limited number of classes. Hence it is desirable to share viewpoint information across classes. Additional complexity arises from unaligned pose labels between classes, i.e. a side view of a car might look more like a frontal view of a toaster, than its side view. To address these problems, we propose a metric learning approach for joint class prediction and pose estimation. Our approach allows to circumvent the problem of viewpoint alignment across multiple classes, and does not require dense viewpoint labels. Moreover, we show, that the learned metric generalizes to new classes, for which the pose labels are not available, and therefore makes it possible to use only partially annotated training sets, relying on the intrinsic similarities in the viewpoint manifolds. We evaluate our approach on two challenging multi-class datasets, 3DObjects and PASCAL3D+.

UR - http://www.scopus.com/inward/record.url?scp=84990043919&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84990043919

T3 - 30th AAAI Conference on Artificial Intelligence, AAAI 2016

SP - 3523

EP - 3529

BT - 30th AAAI Conference on Artificial Intelligence, AAAI 2016

Y2 - 12 February 2016 through 17 February 2016

ER -

Von denselben Autoren