People watching: human actions as a cue for single view geometry

Authors:
David F. Fouhey;Vincent Delaitre;Abhinav Gupta;Alexei A. Efros;Ivan Laptev;Josef Sivic
Affiliations:
Carnegie Mellon University;INRIA/École Normale Supérieure, Paris, France;Carnegie Mellon University;Carnegie Mellon University, USA,INRIA/École Normale Supérieure, Paris, France;INRIA/École Normale Supérieure, Paris, France;INRIA/École Normale Supérieure, Paris, France
Venue:
ECCV'12 Proceedings of the 12th European conference on Computer Vision - Volume Part V
Year:
2012

Citing 20
Cited 1

Histograms of Oriented Gradients for Human Detection

CVPR '05 Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) - Volume 1 - Volume 01
Geometric Context from a Single Image

ICCV '05 Proceedings of the Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1 - Volume 01
Bayesian Autocalibration for Surveillance

ICCV '05 Proceedings of the Tenth IEEE International Conference on Computer Vision - Volume 2
Putting Objects in Perspective

International Journal of Computer Vision
Simultaneous Visual Recognition of Manipulation Actions and Manipulated Objects

ECCV '08 Proceedings of the 10th European Conference on Computer Vision: Part II
Geometric image parsing in man-made environments

ECCV'10 Proceedings of the 11th European conference on Computer vision: Part II
Discriminative learning with latent variables for cluttered indoor scene understanding

ECCV'10 Proceedings of the 11th European conference on Computer vision: Part II
Unsupervised learning of functional categories in video scenes

ECCV'10 Proceedings of the 11th European conference on Computer vision: Part II
Blocks world revisited: image understanding using qualitative geometry and mechanics

ECCV'10 Proceedings of the 11th European conference on Computer vision: Part IV
Thinking inside the box: using appearance models and context based on room geometry

ECCV'10 Proceedings of the 11th European conference on Computer vision: Part VI
Weakly Supervised Learning of Interactions between Humans and Objects

IEEE Transactions on Pattern Analysis and Machine Intelligence
From 3D scene geometry to human workspace

CVPR '11 Proceedings of the 2011 IEEE Conference on Computer Vision and Pattern Recognition
Sampling bedrooms

CVPR '11 Proceedings of the 2011 IEEE Conference on Computer Vision and Pattern Recognition
Learning effective human pose estimation from inaccurate annotation

CVPR '11 Proceedings of the 2011 IEEE Conference on Computer Vision and Pattern Recognition
Functional categorization of objects using real-time markerless motion capture

CVPR '11 Proceedings of the 2011 IEEE Conference on Computer Vision and Pattern Recognition
Scene shape from texture of objects

CVPR '11 Proceedings of the 2011 IEEE Conference on Computer Vision and Pattern Recognition
What makes a chair a chair?

CVPR '11 Proceedings of the 2011 IEEE Conference on Computer Vision and Pattern Recognition
Recovering free space of indoor scenes from a single image

CVPR '12 Proceedings of the 2012 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
Efficient structured prediction for 3D indoor scene understanding

CVPR '12 Proceedings of the 2012 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
Scene semantics from long-term observation of people

ECCV'12 Proceedings of the 12th European conference on Computer Vision - Volume Part VI

Scene semantics from long-term observation of people

ECCV'12 Proceedings of the 12th European conference on Computer Vision - Volume Part VI

Quantified Score

Hi-index	0.00

Visualization

Abstract

We present an approach which exploits the coupling between human actions and scene geometry. We investigate the use of human pose as a cue for single-view 3D scene understanding. Our method builds upon recent advances in still-image pose estimation to extract functional and geometric constraints about the scene. These constraints are then used to improve state-of-the-art single-view 3D scene understanding approaches. The proposed method is validated on a collection of monocular time-lapse sequences collected from YouTube and a dataset of still images of indoor scenes. We demonstrate that observing people performing different actions can significantly improve estimates of 3D scene geometry.