On View Likelihood and Stability
IEEE Transactions on Pattern Analysis and Machine Intelligence
Entity-based aspect graphs: making viewer centered representations more efficient
Pattern Recognition Letters
ACM Computing Surveys (CSUR)
Measuring the performance of shape similarity retrieval methods
Computer Vision and Image Understanding - Special issue on empirical evaluation of computer vision algorithms
Selecting Canonical Views for View-Based 3-D Object Recognition
ICPR '04 Proceedings of the Pattern Recognition, 17th International Conference on (ICPR'04) Volume 2 - Volume 02
The Wisdom of Crowds
A probabilistic approach for 3D shape retrieval by characteristic views
Pattern Recognition Letters
Putting the crowd to work in a knowledge-based factory
Advanced Engineering Informatics
Clustering dictionary definitions using Amazon Mechanical Turk
CSLDAMT '10 Proceedings of the NAACL HLT 2010 Workshop on Creating Speech and Language Data with Amazon's Mechanical Turk
A benchmark for best view selection of 3D objects
Proceedings of the ACM workshop on 3D object retrieval
Fast human classification of 3D object benchmarks
EG 3DOR'10 Proceedings of the 3rd Eurographics conference on 3D Object Retrieval
Hi-index | 0.00 |
The ability to interpret and reason about shapes is a peculiarly human capability that has proven difficult to reproduce algorithmically. So despite the fact that geometric modeling technology has made significant advances in the representation, display and modification of shapes, there have only been incremental advances in geometric reasoning. For example, although today's CAD systems can confidently identify isolated cylindrical holes, they struggle with more ambiguous tasks such as the identification of partial symmetries or similarities in arbitrary geometries. Even well defined problems such as 2D shape nesting or 3D packing generally resist elegant solution and rely instead on brute force explorations of a subset of the many possible solutions. Identifying economic ways to solving such problems would result in significant productivity gains across a wide range of industrial applications. The authors hypothesize that Internet Crowdsourcing might provide a pragmatic way of removing many geometric reasoning bottlenecks. This paper reports the results of experiments conducted with Amazon's mTurk site and designed to determine the feasibility of using Internet Crowdsourcing to carry out geometric reasoning tasks as well as establish some benchmark data for the quality, speed and costs of using this approach. After describing the general architecture and terminology of the mTurk Crowdsourcing system, the paper details the implementation and results of the following three investigations; 1) the identification of "Canonical" viewpoints for individual shapes, 2) the quantification of "similarity" relationships with-in collections of 3D models and 3) the efficient packing of 2D Strips into rectangular areas. The paper concludes with a discussion of the possibilities and limitations of the approach.