AutoHammer: Breaking the Compilation Wall Between Deep Neural Network and Overlay-based FPGA Accelerator.
Kai QianZheng LiuYinqiu LiuHaodong LuZexu ZhangRuiqiu ChenKun WangPublished in: FPGA (2024)
Keyphrases
- neural network
- field programmable gate array
- hardware implementation
- back propagation
- embedded systems
- real time image processing
- artificial neural networks
- hardware architecture
- genetic algorithm
- parallel implementation
- pattern recognition
- high speed
- image reconstruction from projections
- neural nets
- neural network model
- fuzzy artmap
- computing systems
- image processing algorithms
- hardware design
- feed forward neural networks
- backpropagation neural network
- learning vector quantization
- associative memory
- neural network is trained
- signal processing
- network architecture
- training algorithm
- fpga technology
- dedicated hardware
- feedforward neural networks
- low power consumption
- fuzzy logic
- activation function
- parallel computing
- incremental learning