Resilient training of neural network classifiers with approximate computing techniques for hardware-optimised implementations.
Vitor Ferreira TorresFrank Sill TorresPublished in: IET Comput. Digit. Tech. (2019)
Keyphrases
- low cost
- real time
- hardware and software
- hardware implementation
- training samples
- training process
- data structure
- probabilistic model
- efficient computation
- embedded systems
- efficient implementation
- efficiently computing
- software implementation
- parallel architectures
- graphics hardware
- training algorithm
- serious games
- test set
- supervised learning
- object detection
- artificial neural networks
- training data
- case study