BROWSE
Detail
Conference Paper
Modified Spiking Neural Networks for Intelligence Edge
Cited - time in 
Share 
Share
- Authors
- Seongmo Park, Byoung Gun Choi, In Gi Lim, Seongeun Kim, Kyunghwan Park
- Issue Date
- 2020-11
- Citation
- International Conference on Consumer Electronics (ICCE) 2020 : Asia, pp.271-274
- Publisher
- IEEE
- Language
- English
- Type
- Conference Paper
- Abstract
- The proposed algorithm describes MSNN (Modified Spiking Neural Networks) based on hardware for intelligence edge . One way to build an excellent dataset representation is to train Modified Spiking Neural Networks (MSNN) and networks as sensory data for unsupervised tasks. Training various spiking MNIST datasets showed solid evidence that our spike confrontation pairs learn the repetition of the input data from the sensory data to the scene in both the spiking conversion and processing. This algorithm improved a 20% speed up, 93.4% of accuracy and low power operation compared to other networks.
- KSP Keywords
- Speed-up, input data, low-power operation, sensory data, spiking neural networks