Return the end of the valid region for the given dimension @p d */

Flag which specifies if the input tensor has to be reinterpreted as 3D * * @return True if the input tensor has to be reinterpreted as 3D te

Dequantize a scalar value from signed 8-bit asymmetric to floating-point * * @param[in] input Input value to quantize * @param[in] offset Quantiza

Micro-kernel selector * * @param[in] data Selection data passed to help pick the appropriate micro-kernel * * @return A matching micro-kernel else

Dequantize a value given an 8-bit asymmetric quantization scheme * * @param[in] value Value to dequantize * @param[in] scale Scale to use for deq

Set the input and output tensors. * * @param[in] input Source tensor. 3 lower dimensions represent a single input [width, height, IFM]

Depth (third dimension) of the output tensor to be used with the GEMM3D kernel * * @note GEMM3D kernel is used when the output has to be rei

Returns edge id * * @return Edge id */

Convert string values to unsigned long. * * @note This function implements the same behaviour as std::stoul. The latter * is missing in some

Collapses given dimension and above. * * @param[in] dimension Size of the wanted dimension * * @return The linear size of the coll

Memory region size accessor * * @return Memory region size */

Return a copy with collapsed dimensions starting from a given point. * * @param[in] start Starting point of collapsing dimensions. *

Return the kernel to run * * @return The function pointer to the chosen kernel */

Dequantize a neon vector holding 16 16-bit quantized values. * * @param[in] qv Input values to be dequantized. * @param[in] qi Quantization informa

Calculate the Right Hand Side matrix reshaped shape * * @param[in] a Input tensor info * @param[in] rhs_info Right Hand Side matrix informat

Calculate the deep convolution shape output shape of a tensor * * @param[in] input_shape Input tensor shape * @param[in] input_data_layout In

Check if a given data type is of per channel type * * @param[in] dt Input data type. * * @return True if data type is of per channel type, else fa

Natural logorithm. \param arg function argument \return logarithm of \a arg to base e template<typename T> typename enable<expr,T>::type log(T arg) {

Get max sizes. */

GEMMLowp output stage * * @return the GEMMLowp output stage info */

Extract internal representation of a Queue * * @param[in] queue Opaque queue pointer * * @return The internal representation as an IQueue */

Convert a linear index into n-dimensional coordinates. * * @param[in] shape Shape of the n-dimensional tensor. * @param[in] index Linear index spec

Get the pooled height of the layer */

Get the pooled width of the layer */

Calculate the matrix multiplication output shape of two tensors * * @param[in] input0 First input tensor info * @param[in] input

Permutes a given tensor shape given the input and output data layout * * @param[in] tensor_shape Tensor shape to permute * @param[in] in_data_la

Generates appropriate input accessor according to the specified graph parameters * * @param[in] graph_parameters Graph parameters * @param[in] prep