| Class | Description |
|---|---|
| ConvolveDownNoBorderStandard |
Standard implementation of
ConvolveDownNoBorder where no special
optimization has been done. |
| ConvolveDownNoBorderUnrolled_F32_F32 |
Unrolls the convolution kernel to improve runtime performance by reducing array accesses.
|
| ConvolveDownNoBorderUnrolled_S16_I16 |
Unrolls the convolution kernel to improve runtime performance by reducing array accesses.
|
| ConvolveDownNoBorderUnrolled_S16_I16_Div |
Unrolls the convolution kernel to improve runtime performance by reducing array accesses.
|
| ConvolveDownNoBorderUnrolled_U8_I16 |
Unrolls the convolution kernel to improve runtime performance by reducing array accesses.
|
| ConvolveDownNoBorderUnrolled_U8_I8_Div |
Unrolls the convolution kernel to improve runtime performance by reducing array accesses.
|
| ConvolveDownNormalized_JustBorder |
Convolves a 1D kernel in the horizontal or vertical direction while skipping pixels across an image's border.
|
| ConvolveDownNormalizedNaive |
Down convolution with kernel renormalization around image borders.
|
| UtilDownConvolve |