| Autor: |
Hayasaka, Taichi, Nakauchi, Shigeki, Usui, Shiro |
| Předmět: |
|
| Zdroj: |
Systems & Computers in Japan; 1/1/2006, Vol. 37 Issue 1, p105-115, 11p, 7 Graphs |
| Abstrakt: |
Conflicting results in terms of the recognition rate and response times of humans for 3-dimensional object recognition have been obtained on the basis of experiments conducted under various conditions; while some results suggest that these variables vary depending on the viewpoint of an object (viewpoint dependency), others suggest there is no such variation (viewpoint invariance). In this research we have analyzed the issue of viewpoint dependency in object recognition results using support vector machines, a type of machine learning algorithm, based on the belief that 3-dimensional objects are represented in the brain by sets of 2-dimensional data consisting of images of the object from a range of viewpoints. From this perspective it is suggested that the information contained in images of a 3-dimensional object from a range of viewpoints plays a significant role that ties together apparently conflicting results seen in psychophysical experiments regarding the issue of viewpoint dependency or invariance. © 2005 Wiley Periodicals, Inc. Syst Comp Jpn, 37(1): 105–115, 2006; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/scj.10265 [ABSTRACT FROM AUTHOR] |
| Databáze: |
Supplemental Index |
| Externí odkaz: |
|
Plný text