Learning invariant object recognition in the visual system with continuous transformations.
The cerebral cortex utilizes spatiotemporal continuity in the world to help build invariant representations. In vision, these might be representations of objects. The temporal continuity typical of objects has been used in an associative learning rule with a short-term memory trace to help build inv...
| Main Authors: | , , , |
|---|---|
| Format: | Journal article |
| Language: | English |
| Published: |
2006
|
| _version_ | 1797083434916511744 |
|---|---|
| author | Stringer, S Perry, G Rolls, E Proske, J |
| author_facet | Stringer, S Perry, G Rolls, E Proske, J |
| author_sort | Stringer, S |
| collection | OXFORD |
| description | The cerebral cortex utilizes spatiotemporal continuity in the world to help build invariant representations. In vision, these might be representations of objects. The temporal continuity typical of objects has been used in an associative learning rule with a short-term memory trace to help build invariant object representations. In this paper, we show that spatial continuity can also provide a basis for helping a system to self-organize invariant representations. We introduce a new learning paradigm "continuous transformation learning" which operates by mapping spatially similar input patterns to the same postsynaptic neurons in a competitive learning system. As the inputs move through the space of possible continuous transforms (e.g. translation, rotation, etc.), the active synapses are modified onto the set of postsynaptic neurons. Because other transforms of the same stimulus overlap with previously learned exemplars, a common set of postsynaptic neurons is activated by the new transforms, and learning of the new active inputs onto the same postsynaptic neurons is facilitated. We demonstrate that a hierarchical model of cortical processing in the ventral visual system can be trained with continuous transform learning, and highlight differences in the learning of invariant representations to those achieved by trace learning. |
| first_indexed | 2024-03-07T01:41:47Z |
| format | Journal article |
| id | oxford-uuid:971d5c01-c067-4f02-a069-0c4f5e13d833 |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-03-07T01:41:47Z |
| publishDate | 2006 |
| record_format | dspace |
| spelling | oxford-uuid:971d5c01-c067-4f02-a069-0c4f5e13d8332022-03-26T23:57:16ZLearning invariant object recognition in the visual system with continuous transformations.Journal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:971d5c01-c067-4f02-a069-0c4f5e13d833EnglishSymplectic Elements at Oxford2006Stringer, SPerry, GRolls, EProske, JThe cerebral cortex utilizes spatiotemporal continuity in the world to help build invariant representations. In vision, these might be representations of objects. The temporal continuity typical of objects has been used in an associative learning rule with a short-term memory trace to help build invariant object representations. In this paper, we show that spatial continuity can also provide a basis for helping a system to self-organize invariant representations. We introduce a new learning paradigm "continuous transformation learning" which operates by mapping spatially similar input patterns to the same postsynaptic neurons in a competitive learning system. As the inputs move through the space of possible continuous transforms (e.g. translation, rotation, etc.), the active synapses are modified onto the set of postsynaptic neurons. Because other transforms of the same stimulus overlap with previously learned exemplars, a common set of postsynaptic neurons is activated by the new transforms, and learning of the new active inputs onto the same postsynaptic neurons is facilitated. We demonstrate that a hierarchical model of cortical processing in the ventral visual system can be trained with continuous transform learning, and highlight differences in the learning of invariant representations to those achieved by trace learning. |
| spellingShingle | Stringer, S Perry, G Rolls, E Proske, J Learning invariant object recognition in the visual system with continuous transformations. |
| title | Learning invariant object recognition in the visual system with continuous transformations. |
| title_full | Learning invariant object recognition in the visual system with continuous transformations. |
| title_fullStr | Learning invariant object recognition in the visual system with continuous transformations. |
| title_full_unstemmed | Learning invariant object recognition in the visual system with continuous transformations. |
| title_short | Learning invariant object recognition in the visual system with continuous transformations. |
| title_sort | learning invariant object recognition in the visual system with continuous transformations |
| work_keys_str_mv | AT stringers learninginvariantobjectrecognitioninthevisualsystemwithcontinuoustransformations AT perryg learninginvariantobjectrecognitioninthevisualsystemwithcontinuoustransformations AT rollse learninginvariantobjectrecognitioninthevisualsystemwithcontinuoustransformations AT proskej learninginvariantobjectrecognitioninthevisualsystemwithcontinuoustransformations |