Back to Search
Start Over
Is 2D information enough for viewpoint estimation?
- Source :
- Scopus-Elsevier, BMVC
-
Abstract
- © 2014. The copyright of this document resides with its authors. 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 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 viewpoints suppression strategy on the training data lead to comparable or even better performance than 3D methods. Ghodrati A., Pedersoli M., Tuytelaars T., ''Is 2D information enough for viewpoint estimation?'', 25th British machine vision conference - BMVC 2014, 12 pp., September 1-5, 2014, Nottingham, UK. ispartof: pages:1-12 ispartof: Proceedings BMVC 2014 pages:1-12 ispartof: British machine vision conference - BMVC 2014 location:Nottingham, UK. date:1 Sep - 5 Sep 2014 status: published
- Subjects :
- Class (computer programming)
Computer science
business.industry
CAD
PSI_VISICS
Object (computer science)
Viewpoints
Machine learning
computer.software_genre
Convolutional neural network
Minimum bounding box
Encoding (memory)
Artificial intelligence
business
Representation (mathematics)
computer
Subjects
Details
- Database :
- OpenAIRE
- Journal :
- Scopus-Elsevier, BMVC
- Accession number :
- edsair.doi.dedup.....ad5065a4f09db7d015aab9835fb1c50e