TI Deep Learning Product User Guide
sTIDL_PoolingParams_t Struct Reference

Detailed Description

This structure define the parameters spatial Pooling layer in TIDL.

Data Fields

int32_t numChannels
 
int32_t poolingType
 
int32_t kernelW
 
int32_t kernelH
 
int32_t strideW
 
int32_t strideH
 
int32_t padW
 
int32_t padH
 
int32_t inDataQ
 
int32_t outDataQ
 
int32_t useCeil
 

Field Documentation

◆ numChannels

int32_t sTIDL_PoolingParams_t::numChannels

Number of channels channels to be processed

◆ poolingType

int32_t sTIDL_PoolingParams_t::poolingType

Type of the Pooling as defined in eTIDL_PoolType

◆ kernelW

int32_t sTIDL_PoolingParams_t::kernelW

Kernel width

◆ kernelH

int32_t sTIDL_PoolingParams_t::kernelH

Kernel height

◆ strideW

int32_t sTIDL_PoolingParams_t::strideW

Stride in horizontal direction

◆ strideH

int32_t sTIDL_PoolingParams_t::strideH

Stride in vertical direction

◆ padW

int32_t sTIDL_PoolingParams_t::padW

Horizontal Padding requirement in number of elements

◆ padH

int32_t sTIDL_PoolingParams_t::padH

Vertical Padding requirement in number of elements

◆ inDataQ

int32_t sTIDL_PoolingParams_t::inDataQ

Q value of the in data

◆ outDataQ

int32_t sTIDL_PoolingParams_t::outDataQ

Q value expected for out data

◆ useCeil

int32_t sTIDL_PoolingParams_t::useCeil

ceil condition for caffe models