![]() Number of scales must be equal to the number of input dimensions. Constructor & Destructor Documentation ◆ ~IResizeLayer() Warning Do not inherit from this class, as doing so will break forward-compatibility of the API and ABI. ( ResizeCoordinateTransformation::kHALF_PIXEL, ResizeSelector::kUPPER).( ResizeCoordinateTransformation::kHALF_PIXEL, ResizeSelector::kFORMULA).( ResizeCoordinateTransformation::kHALF_PIXEL, ResizeSelector::kFORMULA, ResizeRoundMode::kHALF_UP).( ResizeCoordinateTransformation::kHALF_PIXEL, ResizeSelector::kFORMULA, ResizeRoundMode::kHALF_DOWN).( ResizeCoordinateTransformation::kASYMMETRIC, ResizeSelector::kFORMULA, ResizeRoundMode::kFLOOR).If executing this layer on DLA, the following combinations of parameters are supported: Only static resize layer allows setting scales where the scales are known at build-time. Each output dimension is calculated as floor(input dimension * scale). Dynamic resize layer requires output dimensions to be set as one of the input tensors. Static resize layer requires output dimensions to be known at build-time. It can be done for static as well as dynamic resize layer. The resize layer provides two ways to resize tensor dimensions. The default for all ResizeMode settings (nearest, linear, bilinear, etc.) is ResizeCoordinateTransformation::kASYMMETRIC. The coordinates in the output tensor are mapped to coordinates in the input tensor using a function set by calling setCoordinateTransformation(). ResizeMode::kLINEAR - resizes innermost m dimensions of N-D, where 0 0ĭefault resize mode is ResizeMode::kNEAREST.ResizeMode::kNEAREST - resizes innermost m dimensions of N-D, where 0 0.Resize layer currently supports the following configurations: Resize layer can be used for resizing a N-D tensor. Protected Attributes inherited from nvinfer1::ILayer Protected Member Functions inherited from nvinfer1::INoCopy Protected Member Functions inherited from nvinfer1::ILayer SetMetadata (char const *metadata) noexcept Reset the output type for this layer More. Whether the output type has been set for this layer More. OutputTypeIsSet (int32_t index) const noexcept GetOutputType (int32_t index) const noexcept SetOutputType (int32_t index, DataType dataType) noexcept Reset the computational precision for this layer More. Whether the computational precision has been set for this layer More. Get the computational precision of this layer More. Set the computational precision of this layer. SetPrecision ( DataType dataType) noexcept Replace an input of this layer with a specific tensor. Get the layer output corresponding to the given index. Get the layer input corresponding to the given index. Public Member Functions inherited from nvinfer1::ILayer SetInput (int32_t index, ITensor &tensor) noexceptĪppend or replace an input of this layer with a specific tensor. Get the state for excluding outside pixels. Set the state for excluding outside pixels. SetExcludeOutside (bool excludeFlag) noexcept Get the coefficient 'A' used in cubic interpolation. Set the coefficient 'A' used in cubic interpolation. Get rounding mode for nearest neighbor resize. Set rounding mode for nearest neighbor resize. SetNearestRounding ( ResizeRoundMode value) noexcept Get the coordinate selector function when resized to single pixel. GetSelectorForSinglePixel () const noexcept Set coordinate selector function when resized to single pixel. SetSelectorForSinglePixel ( ResizeSelector selector) noexcept ![]() GetCoordinateTransformation () const noexcept SetCoordinateTransformation ( ResizeCoordinateTransformation coordTransform) noexcept Set whether to align corners while resizing. SetAlignCorners (bool alignCorners) noexcept SetResizeMode ( ResizeMode resizeMode) noexcept GetScales (int32_t size, float *scales) const noexceptĬopies resize scales to scales, where nbScales is the number of scales that were set. SetScales (float const *scales, int32_t nbScales) noexcept ![]() SetOutputDimensions ( Dims dimensions) noexcept ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |