Continuous transformation learning of translation invariant representations.
We show that spatial continuity can enable a network to learn translation invariant representations of objects by self-organization in a hierarchical model of cortical processing in the ventral visual system. During 'continuous transformation learning', the active synapses from each overla...
| Главные авторы: | , , |
|---|---|
| Формат: | Journal article |
| Язык: | English |
| Опубликовано: |
2010
|
| _version_ | 1826274173701324800 |
|---|---|
| author | Perry, G Rolls, E Stringer, S |
| author_facet | Perry, G Rolls, E Stringer, S |
| author_sort | Perry, G |
| collection | OXFORD |
| description | We show that spatial continuity can enable a network to learn translation invariant representations of objects by self-organization in a hierarchical model of cortical processing in the ventral visual system. During 'continuous transformation learning', the active synapses from each overlapping transform are associatively modified onto the set of postsynaptic neurons. Because other transforms of the same object 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 show that the transforms must be close for this to occur; that the temporal order of presentation of each transformed image during training is not crucial for learning to occur; that relatively large numbers of transforms can be learned; and that such continuous transformation learning can be usefully combined with temporal trace training. |
| first_indexed | 2024-03-06T22:39:25Z |
| format | Journal article |
| id | oxford-uuid:5b066a48-8489-4496-969f-6088a7eeef3a |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-03-06T22:39:25Z |
| publishDate | 2010 |
| record_format | dspace |
| spelling | oxford-uuid:5b066a48-8489-4496-969f-6088a7eeef3a2022-03-26T17:19:29ZContinuous transformation learning of translation invariant representations.Journal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:5b066a48-8489-4496-969f-6088a7eeef3aEnglishSymplectic Elements at Oxford2010Perry, GRolls, EStringer, SWe show that spatial continuity can enable a network to learn translation invariant representations of objects by self-organization in a hierarchical model of cortical processing in the ventral visual system. During 'continuous transformation learning', the active synapses from each overlapping transform are associatively modified onto the set of postsynaptic neurons. Because other transforms of the same object 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 show that the transforms must be close for this to occur; that the temporal order of presentation of each transformed image during training is not crucial for learning to occur; that relatively large numbers of transforms can be learned; and that such continuous transformation learning can be usefully combined with temporal trace training. |
| spellingShingle | Perry, G Rolls, E Stringer, S Continuous transformation learning of translation invariant representations. |
| title | Continuous transformation learning of translation invariant representations. |
| title_full | Continuous transformation learning of translation invariant representations. |
| title_fullStr | Continuous transformation learning of translation invariant representations. |
| title_full_unstemmed | Continuous transformation learning of translation invariant representations. |
| title_short | Continuous transformation learning of translation invariant representations. |
| title_sort | continuous transformation learning of translation invariant representations |
| work_keys_str_mv | AT perryg continuoustransformationlearningoftranslationinvariantrepresentations AT rollse continuoustransformationlearningoftranslationinvariantrepresentations AT stringers continuoustransformationlearningoftranslationinvariantrepresentations |