Natural learning of neural networks by reconfiguration
(2003) SPIE Conference on Bioengineered and Bioinspired Systems, 2003 5119. p.273-284- Abstract
- The communicational and computational demands of neural networks are hard to satisfy in a digital technology. Temporal computing solves this problem by iteration, but leaves a slow network. Spatial computing was no option until the coming of modern FPGA devices. The letter shows how a small feed-forward neural module can be configured on the limited logic blocks between RAM and multiplier macro’s. It is then described how by spatial unrolling or by reconfiguration a large modular ANN can be built from such modules.
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/record/603877
- author
- Spaanenburg, Lambert LU ; Alberts, R ; Slump, C H and vanderZwaag, B J
- organization
- publishing date
- 2003
- type
- Chapter in Book/Report/Conference proceeding
- publication status
- published
- subject
- keywords
- Spatial Computing, ANN, Modularity, FPGA, Reconfiguration
- host publication
- SPIE Proceedings on Bioengineered and Bioinspired Systems
- volume
- 5119
- pages
- 273 - 284
- publisher
- SPIE
- conference name
- SPIE Conference on Bioengineered and Bioinspired Systems, 2003
- conference location
- Maspalomas, Gran Canaria, Spain
- conference dates
- 2003-05-19
- external identifiers
-
- wos:000183951000031
- scopus:0041827012
- ISSN
- 1996-756X
- 0277-786X
- DOI
- 10.1117/12.499549
- language
- English
- LU publication?
- yes
- id
- c9d0b592-fba2-437c-82c6-e60365f7171f (old id 603877)
- date added to LUP
- 2016-04-01 12:05:00
- date last changed
- 2024-01-08 07:44:31
@inproceedings{c9d0b592-fba2-437c-82c6-e60365f7171f, abstract = {{The communicational and computational demands of neural networks are hard to satisfy in a digital technology. Temporal computing solves this problem by iteration, but leaves a slow network. Spatial computing was no option until the coming of modern FPGA devices. The letter shows how a small feed-forward neural module can be configured on the limited logic blocks between RAM and multiplier macro’s. It is then described how by spatial unrolling or by reconfiguration a large modular ANN can be built from such modules.}}, author = {{Spaanenburg, Lambert and Alberts, R and Slump, C H and vanderZwaag, B J}}, booktitle = {{SPIE Proceedings on Bioengineered and Bioinspired Systems}}, issn = {{1996-756X}}, keywords = {{Spatial Computing; ANN; Modularity; FPGA; Reconfiguration}}, language = {{eng}}, pages = {{273--284}}, publisher = {{SPIE}}, title = {{Natural learning of neural networks by reconfiguration}}, url = {{http://dx.doi.org/10.1117/12.499549}}, doi = {{10.1117/12.499549}}, volume = {{5119}}, year = {{2003}}, }