Low-power implementation of Mitchell's approximate logarithmic multiplication for convolutional neural networks.
Min Soo KimAlberto A. Del BarrioRomán HermidaNader BagherzadehPublished in: ASP-DAC (2018)
Keyphrases
- low power
- power consumption
- low cost
- high speed
- convolutional neural networks
- vlsi architecture
- cmos technology
- single chip
- signal processor
- high power
- low power consumption
- vlsi circuits
- ultra low power
- logic circuits
- efficient implementation
- wireless transmission
- hardware and software
- design methodology
- image sensor
- general purpose
- digital signal processing
- embedded systems
- low complexity