Detecting humans via their pose

Alessandro Bissacco, Ming Hsuan Yang, Stefano Soatto

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

29 Citations (Scopus)

Abstract

We consider the problem of detecting humans and classifying their pose from a single image. Specifically, our goal is to devise a statistical model that simultaneously answers two questions: 1) is there a human in the image? and, if so, 2) what is a low-dimensional representation of her pose? We investigate models that can be learned in an unsupervised manner on unlabeled images of human poses, and provide information that can be used to match the pose of a new image to the ones present in the training set. Starting from a set of descriptors recently proposed for human detection, we apply the Latent Dirichlet Allocation framework to model the statistics of these features, and use the resulting model to answer the above questions. We show how our model can efficiently describe the space of images of humans with their pose, by providing an effective representation of poses for tasks such as classification and matching, while performing remarkably well in human/non human decision problems, thus enabling its use for human detection. We validate the model with extensive quantitative experiments and comparisons with other approaches on human detection and pose matching.

Original languageEnglish
Title of host publicationAdvances in Neural Information Processing Systems 19 - Proceedings of the 2006 Conference
Pages169-176
Number of pages8
Publication statusPublished - 2007 Dec 1
Event20th Annual Conference on Neural Information Processing Systems, NIPS 2006 - Vancouver, BC, Canada
Duration: 2006 Dec 42006 Dec 7

Publication series

NameAdvances in Neural Information Processing Systems
ISSN (Print)1049-5258

Conference

Conference20th Annual Conference on Neural Information Processing Systems, NIPS 2006
CountryCanada
CityVancouver, BC
Period06/12/406/12/7

Fingerprint

Statistics
Experiments
Statistical Models

All Science Journal Classification (ASJC) codes

  • Computer Networks and Communications
  • Information Systems
  • Signal Processing

Cite this

Bissacco, A., Yang, M. H., & Soatto, S. (2007). Detecting humans via their pose. In Advances in Neural Information Processing Systems 19 - Proceedings of the 2006 Conference (pp. 169-176). (Advances in Neural Information Processing Systems).
Bissacco, Alessandro ; Yang, Ming Hsuan ; Soatto, Stefano. / Detecting humans via their pose. Advances in Neural Information Processing Systems 19 - Proceedings of the 2006 Conference. 2007. pp. 169-176 (Advances in Neural Information Processing Systems).
@inproceedings{ef20155e23d04588b66f490f3d015c16,
title = "Detecting humans via their pose",
abstract = "We consider the problem of detecting humans and classifying their pose from a single image. Specifically, our goal is to devise a statistical model that simultaneously answers two questions: 1) is there a human in the image? and, if so, 2) what is a low-dimensional representation of her pose? We investigate models that can be learned in an unsupervised manner on unlabeled images of human poses, and provide information that can be used to match the pose of a new image to the ones present in the training set. Starting from a set of descriptors recently proposed for human detection, we apply the Latent Dirichlet Allocation framework to model the statistics of these features, and use the resulting model to answer the above questions. We show how our model can efficiently describe the space of images of humans with their pose, by providing an effective representation of poses for tasks such as classification and matching, while performing remarkably well in human/non human decision problems, thus enabling its use for human detection. We validate the model with extensive quantitative experiments and comparisons with other approaches on human detection and pose matching.",
author = "Alessandro Bissacco and Yang, {Ming Hsuan} and Stefano Soatto",
year = "2007",
month = "12",
day = "1",
language = "English",
isbn = "9780262195683",
series = "Advances in Neural Information Processing Systems",
pages = "169--176",
booktitle = "Advances in Neural Information Processing Systems 19 - Proceedings of the 2006 Conference",

}

Bissacco, A, Yang, MH & Soatto, S 2007, Detecting humans via their pose. in Advances in Neural Information Processing Systems 19 - Proceedings of the 2006 Conference. Advances in Neural Information Processing Systems, pp. 169-176, 20th Annual Conference on Neural Information Processing Systems, NIPS 2006, Vancouver, BC, Canada, 06/12/4.

Detecting humans via their pose. / Bissacco, Alessandro; Yang, Ming Hsuan; Soatto, Stefano.

Advances in Neural Information Processing Systems 19 - Proceedings of the 2006 Conference. 2007. p. 169-176 (Advances in Neural Information Processing Systems).

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

TY - GEN

T1 - Detecting humans via their pose

AU - Bissacco, Alessandro

AU - Yang, Ming Hsuan

AU - Soatto, Stefano

PY - 2007/12/1

Y1 - 2007/12/1

N2 - We consider the problem of detecting humans and classifying their pose from a single image. Specifically, our goal is to devise a statistical model that simultaneously answers two questions: 1) is there a human in the image? and, if so, 2) what is a low-dimensional representation of her pose? We investigate models that can be learned in an unsupervised manner on unlabeled images of human poses, and provide information that can be used to match the pose of a new image to the ones present in the training set. Starting from a set of descriptors recently proposed for human detection, we apply the Latent Dirichlet Allocation framework to model the statistics of these features, and use the resulting model to answer the above questions. We show how our model can efficiently describe the space of images of humans with their pose, by providing an effective representation of poses for tasks such as classification and matching, while performing remarkably well in human/non human decision problems, thus enabling its use for human detection. We validate the model with extensive quantitative experiments and comparisons with other approaches on human detection and pose matching.

AB - We consider the problem of detecting humans and classifying their pose from a single image. Specifically, our goal is to devise a statistical model that simultaneously answers two questions: 1) is there a human in the image? and, if so, 2) what is a low-dimensional representation of her pose? We investigate models that can be learned in an unsupervised manner on unlabeled images of human poses, and provide information that can be used to match the pose of a new image to the ones present in the training set. Starting from a set of descriptors recently proposed for human detection, we apply the Latent Dirichlet Allocation framework to model the statistics of these features, and use the resulting model to answer the above questions. We show how our model can efficiently describe the space of images of humans with their pose, by providing an effective representation of poses for tasks such as classification and matching, while performing remarkably well in human/non human decision problems, thus enabling its use for human detection. We validate the model with extensive quantitative experiments and comparisons with other approaches on human detection and pose matching.

UR - http://www.scopus.com/inward/record.url?scp=69549126943&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=69549126943&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:69549126943

SN - 9780262195683

T3 - Advances in Neural Information Processing Systems

SP - 169

EP - 176

BT - Advances in Neural Information Processing Systems 19 - Proceedings of the 2006 Conference

ER -

Bissacco A, Yang MH, Soatto S. Detecting humans via their pose. In Advances in Neural Information Processing Systems 19 - Proceedings of the 2006 Conference. 2007. p. 169-176. (Advances in Neural Information Processing Systems).