An FPGA-based floating-point processor array supporting a high-precision dot product.
Fritz Mayer-LindenbergValerij BellerPublished in: FPT (2006)
Keyphrases
- high precision
- floating point
- dot product
- processor array
- high recall
- fixed point
- kernel function
- similarity function
- parallel algorithm
- parallel implementation
- floating point arithmetic
- mesh connected
- instruction set
- gaussian kernels
- feature space
- hardware implementation
- parallel computers
- computer vision
- information extraction
- preprocessing
- floating point unit