A Self-Organizing Multiple-View Representation of 3D Objects
We explore representation of 3D objects in which several distinct 2D views are stored for each object. We demonstrate the ability of a two-layer network of thresholded summation units to support such representations. Using unsupervised Hebbian relaxation, we trained the network to recognise ten obje...
| Main Authors: | , |
|---|---|
| Language: | en_US |
| Published: |
2004
|
| Online Access: | http://hdl.handle.net/1721.1/6514 |
| _version_ | 1826190390906060800 |
|---|---|
| author | Edelman, Shimon Weinshall, Daphna |
| author_facet | Edelman, Shimon Weinshall, Daphna |
| author_sort | Edelman, Shimon |
| collection | MIT |
| description | We explore representation of 3D objects in which several distinct 2D views are stored for each object. We demonstrate the ability of a two-layer network of thresholded summation units to support such representations. Using unsupervised Hebbian relaxation, we trained the network to recognise ten objects from different viewpoints. The training process led to the emergence of compact representations of the specific input views. When tested on novel views of the same objects, the network exhibited a substantial generalisation capability. In simulated psychophysical experiments, the network's behavior was qualitatively similar to that of human subjects. |
| first_indexed | 2024-09-23T08:39:34Z |
| id | mit-1721.1/6514 |
| institution | Massachusetts Institute of Technology |
| language | en_US |
| last_indexed | 2024-09-23T08:39:34Z |
| publishDate | 2004 |
| record_format | dspace |
| spelling | mit-1721.1/65142019-04-09T19:01:55Z A Self-Organizing Multiple-View Representation of 3D Objects Edelman, Shimon Weinshall, Daphna We explore representation of 3D objects in which several distinct 2D views are stored for each object. We demonstrate the ability of a two-layer network of thresholded summation units to support such representations. Using unsupervised Hebbian relaxation, we trained the network to recognise ten objects from different viewpoints. The training process led to the emergence of compact representations of the specific input views. When tested on novel views of the same objects, the network exhibited a substantial generalisation capability. In simulated psychophysical experiments, the network's behavior was qualitatively similar to that of human subjects. 2004-10-04T15:14:24Z 2004-10-04T15:14:24Z 1989-08-01 AIM-1146 http://hdl.handle.net/1721.1/6514 en_US AIM-1146 2399506 bytes 1875063 bytes application/postscript application/pdf application/postscript application/pdf |
| spellingShingle | Edelman, Shimon Weinshall, Daphna A Self-Organizing Multiple-View Representation of 3D Objects |
| title | A Self-Organizing Multiple-View Representation of 3D Objects |
| title_full | A Self-Organizing Multiple-View Representation of 3D Objects |
| title_fullStr | A Self-Organizing Multiple-View Representation of 3D Objects |
| title_full_unstemmed | A Self-Organizing Multiple-View Representation of 3D Objects |
| title_short | A Self-Organizing Multiple-View Representation of 3D Objects |
| title_sort | self organizing multiple view representation of 3d objects |
| url | http://hdl.handle.net/1721.1/6514 |
| work_keys_str_mv | AT edelmanshimon aselforganizingmultipleviewrepresentationof3dobjects AT weinshalldaphna aselforganizingmultipleviewrepresentationof3dobjects AT edelmanshimon selforganizingmultipleviewrepresentationof3dobjects AT weinshalldaphna selforganizingmultipleviewrepresentationof3dobjects |