Analog CMOS-based resistive processing unit for deep neural network training.
Seyoung KimTayfun GokmenHyung-Min LeeWilfried E. HaenschPublished in: MWSCAS (2017)
Keyphrases
- neural network training
- processing units
- analog vlsi
- cmos image sensor
- parallel processing
- circuit design
- neural network
- training algorithm
- focal plane
- parallel computing
- computing systems
- optimization method
- floating gate
- dynamic range
- power consumption
- machine learning
- particle swarm optimisation
- analog to digital converter
- data processing
- real world
- genetic algorithm
- image segmentation
- feature vectors
- data sets
- low power
- general purpose