Application of Neuromorphic Devices Using Stacked IGZO Thin Film to Intellectual Learning
-
- IWAGI Etsuko
- Ryukoku University
-
- KAWANISHI Hidenori
- Ryukoku University
-
- KIMURA Mutsumi
- Ryukoku University,Nara Institute of Science and Technology (NAIST)
Bibliographic Information
- Other Title
-
- 積層In-Ga-Zn-O薄膜を利用したニューロモルフィックデバイスの知的学習への応用
Abstract
We have developed double layer neuromorphic devices using oxide semiconductors. Plastic properties of amorphous In-Ga-Zn-O were used as artificial synapses in a crooss-bar configuration. We have combined this artificial synapse with artificial neurons formed in FPGA in order to test its performance in pattern recognition. We have successfully demonstrated the learning process of two letter patterns by introducing modified Hebb's rule.which is suitable for electronic artificial neural networks. They are in good agreement with the simulation results. This results shows that this approach is quite useful as neuromorphic devices.
- Tweet
Details 詳細情報について
-
- CRID
- 1390014104843263232
-
- ISSN
- 18810217
-
- Text Lang
- ja
-
- Data Source
-
- JaLC
- KAKEN
-
- Abstract License Flag
- Disallowed