NeuroBlend: Towards Low-Power yet Accurate Neural Network-Based Inference Engine Blending Binary and Fixed-Point Convolutions.
Arash FayyaziMahdi NazemiArya FayyaziMassoud PedramPublished in: ACM Great Lakes Symposium on VLSI (2024)
Keyphrases
- fixed point
- low power
- inference engine
- power consumption
- low cost
- high speed
- rule base
- knowledge base
- expert systems
- knowledge based systems
- knowledge representation
- sufficient conditions
- logic circuits
- neural network
- low power consumption
- cmos technology
- fixed point theorem
- dynamical systems
- belief propagation
- floating point
- high quality
- gate array
- machine learning
- high resolution
- association rules
- image segmentation
- image processing
- real time