MMALIB User Guide
MMALIB_CNN_convolve_col_smallNo_highPrecision_pointwisePost_reorderWeights.h File Reference

Go to the source code of this file.

Data Structures

struct  MMALIB_CNN_convolve_col_smallNo_highPrecision_pointwisePost_reorderWeights_Args
 This structure holds all the input parameters for reordering CNN filter weights for column-based convolution kernel. More...
 

Enumerations

enum  MMALIB_reorder_weights_highPrecision_pointwisePost_operation_e { HIGHPRECISION_POINTWISEPOST_REORDERWEIGHTS_AND_BIAS , HIGHPRECISION_POINTWISEPOST_REORDERWEIGHTS , HIGHPRECISION_POINTWISEPOST_REORDERBIAS }
 Enumeration for different operations for the reorderWeights function. More...
 

Functions

int32_t MMALIB_CNN_convolve_col_smallNo_highPrecision_pointwisePost_reorderWeights_getMemorySize (const MMALIB_CNN_convolve_col_smallNo_highPrecision_pointwisePost_reorderWeights_Args *pArgs)
 This function returns the amount of memory that needs to be allocated for reordered kernel coefficients needed to support MMALIB_CNN_convolve_col_smallNo_highPrecision. More...
 
MMALIB_STATUS MMALIB_CNN_convolve_col_smallNo_highPrecision_pointwisePost_reorderWeights_fillBufParams (const MMALIB_CNN_convolve_col_smallNo_highPrecision_pointwisePost_reorderWeights_Args *pArgs, MMALIB_bufParams3D_t *pReorderedWeights_addr)
 This function populates a structure of dimensional information about the reordered coefficients buffer from a MMALIB_CNN_convolve_col_smallNo_highPrecision_pointwisePost_reorderWeights_Args structure. More...
 
MMALIB_STATUS MMALIB_CNN_convolve_col_smallNo_highPrecision_pointwisePost_reorderWeights_exec (uint32_t operation, const MMALIB_CNN_convolve_col_smallNo_highPrecision_pointwisePost_reorderWeights_Args *pArgs, const MMALIB_bufParams3D_t *pWeights_addr, const void *restrict pWeights, const MMALIB_bufParams2D_t *pBias_addr, const void *restrict pBias, const MMALIB_bufParams3D_t *pReorderedWeights_addr, void *restrict pReorderedWeights)
 This function takes a set of weights and biases (optional) and reorders the coefficients for improved efficiency when computing column flow convolution. More...