Features for Recognition: Viewpoint Invariance for Non-Planar Scenes

Authors:
Andrea Vedaldi;Stefano Soatto
Affiliations:
University of California at Los Angeles;University of California at Los Angeles
Venue:
ICCV '05 Proceedings of the Tenth IEEE International Conference on Computer Vision - Volume 2
Year:
2005

Citing 0
Cited 10

Mumford-Shah on the Move: Region-Based Segmentation on Deforming Manifolds with Application to 3-D Reconstruction of Shape and Appearance from Multi-View Images

Journal of Mathematical Imaging and Vision
Local invariant feature detectors: a survey

Foundations and Trends® in Computer Graphics and Vision
Localizing Objects with Smart Dictionaries

ECCV '08 Proceedings of the 10th European Conference on Computer Vision: Part I
Balancing deformability and discriminability for shape matching

ECCV'10 Proceedings of the 11th European conference on computer vision conference on Computer vision: Part III
Visual-inertial navigation, mapping and localization: A scalable real-time causal approach

International Journal of Robotics Research
Wide-baseline correspondence from locally affine invariant contour matching

ICIAR'11 Proceedings of the 8th international conference on Image analysis and recognition - Volume Part I
Viewpoint induced deformation statistics and the design of viewpoint invariant features: singularities and occlusions

ECCV'06 Proceedings of the 9th European conference on Computer Vision - Volume Part II
Self-similar sketch

ECCV'12 Proceedings of the 12th European conference on Computer Vision - Volume Part II
A Multisensor Architecture Providing Location-based Services for Smartphones

Mobile Networks and Applications
Special Section on Computer Graphics in Brazil: Invariances of single curved manifolds applied to mesh segmentation

Computers and Graphics

Quantified Score

Hi-index	0.00

Visualization

Abstract

Most current local feature detectors/descriptors implicitly assume that the scene is (locally) planar, an assumption that is violated at surface discontinuities. We show that this restriction is, at least in theory, unnecessary, as one can construct local features that are viewpoint-invariant for generic non-planar scenes. However, we show that any such feature necessarily sacrifices shape information, in the sense of being non shape-discriminative. Finally, we show that if viewpoint is factored out as part of the matching process, rather than explicitly in the representation, then shape is discriminative indeed. We illustrate our theoretical results empirically by showing that, even for simple scenes, current affine descriptors fail where even a naive 3-D viewpoint invariant succeeds in matching.