RPPT Tensor Operations - Filter Augmentations.

RPPT Tensor Operations - Filter Augmentations.#

RPP: RPPT Tensor Operations - Filter Augmentations.

RPPT Tensor Operations - Filter Augmentations. More...

Functions
RppStatus	rppt_box_filter_gpu (RppPtr_t srcPtr, RpptDescPtr srcDescPtr, RppPtr_t dstPtr, RpptDescPtr dstDescPtr, Rpp32u kernelSize, RpptROIPtr roiTensorPtrSrc, RpptRoiType roiType, rppHandle_t rppHandle)
	Box Filter augmentation on HIP backend for a NCHW/NHWC layout tensor. More...

RppStatus	rppt_gaussian_filter_gpu (RppPtr_t srcPtr, RpptDescPtr srcDescPtr, RppPtr_t dstPtr, RpptDescPtr dstDescPtr, Rpp32f *stdDevTensor, Rpp32u kernelSize, RpptROIPtr roiTensorPtrSrc, RpptRoiType roiType, rppHandle_t rppHandle)
	Gaussian Filter augmentation on HIP backend for a NCHW/NHWC layout tensor. More...

RppStatus rppt_box_filter_gpu	(	RppPtr_t	srcPtr,
		RpptDescPtr	srcDescPtr,
		RppPtr_t	dstPtr,
		RpptDescPtr	dstDescPtr,
		Rpp32u	kernelSize,
		RpptROIPtr	roiTensorPtrSrc,
		RpptRoiType	roiType,
		rppHandle_t	rppHandle
	)

Box Filter augmentation on HIP backend for a NCHW/NHWC layout tensor.

The box filter augmentation runs for a batch of RGB(3 channel) / greyscale(1 channel) images with an NHWC/NCHW tensor layout.

srcPtr depth ranges - Rpp8u (0 to 255), Rpp16f (0 to 1), Rpp32f (0 to 1), Rpp8s (-128 to 127).

dstPtr depth ranges - Will be same depth as srcPtr.

Sample Input

Sample 3x3 Output

Sample 5x5 Output

Sample 7x7 Output

Sample 9x9 Output

Parameters

[in]	srcPtr	source tensor in HIP memory
[in]	srcDescPtr	source tensor descriptor (Restrictions - numDims = 4, offsetInBytes >= 0, dataType = U8/F16/F32/I8, layout = NCHW/NHWC, c = 1/3)
[out]	dstPtr	destination tensor in HIP memory
[in]	dstDescPtr	destination tensor descriptor (Restrictions - numDims = 4, offsetInBytes >= 0, dataType = U8/F16/F32/I8, layout = NCHW/NHWC, c = same as that of srcDescPtr)
[in]	kernelSize	kernel size for box filter (a single Rpp32u odd number with kernelSize = 3/5/7/9 that applies to all images in the batch)
[in]	roiTensorSrc	ROI data in HIP memory, for each image in source tensor (2D tensor of size batchSize * 4, in either format - XYWH(xy.x, xy.y, roiWidth, roiHeight) or LTRB(lt.x, lt.y, rb.x, rb.y))
[in]	roiType	ROI type used (RpptRoiType::XYWH or RpptRoiType::LTRB)
[in]	rppHandle	RPP HIP handle created with `rppCreateWithStreamAndBatchSize()`

Return values

RPP_SUCCESS	Successful completion.
RPP_ERROR*	Unsuccessful completion.

RppStatus rppt_gaussian_filter_gpu	(	RppPtr_t	srcPtr,
		RpptDescPtr	srcDescPtr,
		RppPtr_t	dstPtr,
		RpptDescPtr	dstDescPtr,
		Rpp32f *	stdDevTensor,
		Rpp32u	kernelSize,
		RpptROIPtr	roiTensorPtrSrc,
		RpptRoiType	roiType,
		rppHandle_t	rppHandle
	)

Gaussian Filter augmentation on HIP backend for a NCHW/NHWC layout tensor.

The gaussian filter augmentation runs for a batch of RGB(3 channel) / greyscale(1 channel) images with an NHWC/NCHW tensor layout.

srcPtr depth ranges - Rpp8u (0 to 255), Rpp16f (0 to 1), Rpp32f (0 to 1), Rpp8s (-128 to 127).