FFConv: An FPGA-based Accelerator for Fast Convolution Layers in Convolutional Neural Networks.
Afzal AhmadMuhammad Adeel PashaPublished in: ACM Trans. Embed. Comput. Syst. (2020)
Keyphrases
- convolutional neural networks
- field programmable gate array
- convolutional network
- hardware implementation
- image processing
- multi layer
- video processing
- convolution kernel
- application specific
- parallel implementation
- embedded systems
- data sets
- motion estimation
- gis data
- multiple layers
- hardware design
- neural network
- data mining
- multiscale
- compute intensive
- single layer
- hardware architecture
- fast fourier transform
- fourier domain
- parallel computing
- coarse to fine