Reducing Memory Requirements of Convolutional Neural Networks for Inference at the Edge.
Tomas BravenecTomas FryzaPublished in: RADIOELEKTRONIKA (2021)
Keyphrases
- memory requirements
- convolutional neural networks
- memory space
- convolutional network
- computational complexity
- memory usage
- bayesian networks
- edge detection
- bayesian inference
- computational power
- inference process
- low memory
- probabilistic inference
- edge information
- computational speed
- grammatical inference
- belief networks
- undirected graph
- edge map
- edge detector
- dynamic programming
- data sets