Is 2D Information Enough For Viewpoint Estimation?

Amir Ghodrati, Marco Pedersoli and Tinne Tuytelaars

In Proceedings British Machine Vision Conference 2014
http://dx.doi.org/10.5244/C.28.19

Abstract

Recent top performing methods for viewpoint estimation make use of 3D information like 3D CAD models or 3D landmarks to build a 3D representation of the class. These 3D annotations are expensive and not really available for many classes. In this paper we investigate whether and how comparable or better performance can be obtained without any 3D information. We consider viewpoint estimation as a 1-vs-all classification problem on the previously detected object bounding box. In this framework we compare several features and parameter configurations and show that the modern representations based on Fisher encoding and convolutional neural network based features together with a neighbor viewpoint suppression strategy on the training data lead to comparable or better performance than the more annotation demanding 3D methods.

Session

3D and Stereo

Files

Extended Abstract (PDF, 1 page, 118K)
Paper (PDF, 12 pages, 319K)
Bibtex File

Presentation

Citation

Amir Ghodrati, Marco Pedersoli, and Tinne Tuytelaars. Is 2D Information Enough For Viewpoint Estimation?. Proceedings of the British Machine Vision Conference. BMVA Press, September 2014.

BibTex

@inproceedings{BMVC.28.19
	title = {Is 2D Information Enough For Viewpoint Estimation?},
	author = {Ghodrati, Amir and Pedersoli, Marco and Tuytelaars, Tinne},
	year = {2014},
	booktitle = {Proceedings of the British Machine Vision Conference},
	publisher = {BMVA Press},
	editors = {Valstar, Michel and French, Andrew and Pridmore, Tony}
	doi = { http://dx.doi.org/10.5244/C.28.19 }
}