Fused Depthwise Tiling for Memory Optimization in TinyML Deep Neural Network Inference.
Rafael StahlDaniel Mueller-GritschnederUlf SchlichtmannPublished in: CoRR (2023)
Keyphrases
- neural network
- auto associative
- image reconstruction from projections
- data fusion
- back propagation
- limited memory
- neural nets
- associative memory
- neural network model
- optimization problems
- radial basis function
- computing power
- recurrent neural networks
- optimization process
- artificial neural networks
- pattern recognition
- highly non linear
- genetic algorithm
- bayesian networks
- data sets
- memory space
- training algorithm
- probabilistic inference
- multilayer perceptron
- memory requirements
- fault diagnosis
- optimization algorithm
- optimization methods
- network model
- belief networks
- training process
- information fusion
- network architecture
- memory usage
- random access
- global optimization
- memory size
- self organizing maps
- dual decomposition