Visual Recognition and Categorization on the Basis of Similarities to Multiple Class Prototypes
To recognize a previously seen object, the visual system must overcome the variability in the object's appearance caused by factors such as illumination and pose. Developments in computer vision suggest that it may be possible to counter the influence of these factors, by learning to inte...
| Main Authors: | , |
|---|---|
| Language: | en_US |
| Published: |
2004
|
| Online Access: | http://hdl.handle.net/1721.1/7248 |
| _version_ | 1826191960257331200 |
|---|---|
| author | Edelman, Shimon Duvdevani-Bar, Sharon |
| author_facet | Edelman, Shimon Duvdevani-Bar, Sharon |
| author_sort | Edelman, Shimon |
| collection | MIT |
| description | To recognize a previously seen object, the visual system must overcome the variability in the object's appearance caused by factors such as illumination and pose. Developments in computer vision suggest that it may be possible to counter the influence of these factors, by learning to interpolate between stored views of the target object, taken under representative combinations of viewing conditions. Daily life situations, however, typically require categorization, rather than recognition, of objects. Due to the open-ended character both of natural kinds and of artificial categories, categorization cannot rely on interpolation between stored examples. Nonetheless, knowledge of several representative members, or prototypes, of each of the categories of interest can still provide the necessary computational substrate for the categorization of new instances. The resulting representational scheme based on similarities to prototypes appears to be computationally viable, and is readily mapped onto the mechanisms of biological vision revealed by recent psychophysical and physiological studies. |
| first_indexed | 2024-09-23T09:04:01Z |
| id | mit-1721.1/7248 |
| institution | Massachusetts Institute of Technology |
| language | en_US |
| last_indexed | 2024-09-23T09:04:01Z |
| publishDate | 2004 |
| record_format | dspace |
| spelling | mit-1721.1/72482019-04-10T20:30:48Z Visual Recognition and Categorization on the Basis of Similarities to Multiple Class Prototypes Edelman, Shimon Duvdevani-Bar, Sharon To recognize a previously seen object, the visual system must overcome the variability in the object's appearance caused by factors such as illumination and pose. Developments in computer vision suggest that it may be possible to counter the influence of these factors, by learning to interpolate between stored views of the target object, taken under representative combinations of viewing conditions. Daily life situations, however, typically require categorization, rather than recognition, of objects. Due to the open-ended character both of natural kinds and of artificial categories, categorization cannot rely on interpolation between stored examples. Nonetheless, knowledge of several representative members, or prototypes, of each of the categories of interest can still provide the necessary computational substrate for the categorization of new instances. The resulting representational scheme based on similarities to prototypes appears to be computationally viable, and is readily mapped onto the mechanisms of biological vision revealed by recent psychophysical and physiological studies. 2004-10-20T21:04:09Z 2004-10-20T21:04:09Z 1997-09-01 AIM-1615 CBCL-154 http://hdl.handle.net/1721.1/7248 en_US AIM-1615 CBCL-154 3735406 bytes 1301633 bytes application/postscript application/pdf application/postscript application/pdf |
| spellingShingle | Edelman, Shimon Duvdevani-Bar, Sharon Visual Recognition and Categorization on the Basis of Similarities to Multiple Class Prototypes |
| title | Visual Recognition and Categorization on the Basis of Similarities to Multiple Class Prototypes |
| title_full | Visual Recognition and Categorization on the Basis of Similarities to Multiple Class Prototypes |
| title_fullStr | Visual Recognition and Categorization on the Basis of Similarities to Multiple Class Prototypes |
| title_full_unstemmed | Visual Recognition and Categorization on the Basis of Similarities to Multiple Class Prototypes |
| title_short | Visual Recognition and Categorization on the Basis of Similarities to Multiple Class Prototypes |
| title_sort | visual recognition and categorization on the basis of similarities to multiple class prototypes |
| url | http://hdl.handle.net/1721.1/7248 |
| work_keys_str_mv | AT edelmanshimon visualrecognitionandcategorizationonthebasisofsimilaritiestomultipleclassprototypes AT duvdevanibarsharon visualrecognitionandcategorizationonthebasisofsimilaritiestomultipleclassprototypes |