* @brief Get the shape of the sample. * * [shape_access] */

* @brief Transforms DALI's OpenCv-style bayer pattern specification to NPP's enum. * All the supported bayer patterns are 2x2 tiles. * For example t

Converts to torch.Tensor, either copying the data or using dlpack

Mark `if` (true) branch as current scope.

Code for map calculation. Based on https://github.com/opencv/opencv/blob/3.4/samples/python/tutorial_code/ImgTrans/remap/Remap_Demo.py # noq

Pickles `message` instances, stores it in the provided `shm` chunk at given offset and returns `ShmMessageDesc` instance describing the place

* @brief Boilerplate code for defining inputs to the graph (input and pred nodes). */

* @brief Calculates the shape of the result of reduction under given parameters * * The reduced shape is calculated in one of two ways: * 1. If `ke

A workaround for category comparison acros shared object boundary

* @brief Verifies that the inputs in the workspace satisfy the layout * constraints imposed by the schema. */

* @brief Check if the ArgumentInput of given shape can be used with GetArgument(), * representing a batch of scalars * * @argument sho

* @brief Clears the contents of the workspaces, reseting it * to a default state. */

* @brief Reset the outputs held in the workspace, releasing the current memory and setting the * effective batch size to 0 for every output. */

* @brief Sorts and merges contiguous ranges */