Primitives
Collaboration diagram for Primitives:
|
Classes | |
| struct | cldnn::activation |
| Activation using rectified linear unit or parameterized rectified linear unit. More... | |
| struct | cldnn::batch_norm |
| Batch normalization primitive. More... | |
| struct | cldnn::concatenation |
| struct | cldnn::convolution |
| Performs forward spatial convolution with weight sharing. Also supports built-in Relu cldnn_activation_desc available by setting it in arguments. More... | |
| struct | cldnn::crop |
| Performs crop operation on input. More... | |
| struct | cldnn::custom_gpu_primitive |
| This primitive executes a custom kernel provided by the application. More... | |
| struct | cldnn::data |
| Provides input data to topology. More... | |
| struct | cldnn::deconvolution |
| Performs transposed convolution. Also supports built-in Relu activation available by setting it in arguments. More... | |
| struct | cldnn::detection_output |
| Generates a list of detections based on location and confidence predictions by doing non maximum suppression. More... | |
| struct | cldnn::eltwise |
| Performs elementwise operations (sum, subtract, max or product) on two input primitives Also supports built-in Relu activation available by setting it in arguments. . More... | |
| struct | cldnn::fully_connected |
| Performs forward fully connected layer (inner product). Also supports built-in Relu cldnn_activation_desc available by setting it in arguments. . More... | |
| struct | cldnn::input_layout |
| Provides input layout for a data to be passed later to network. More... | |
| struct | cldnn::lrn |
| Local response normalization. More... | |
| struct | cldnn::normalize |
| Normalizes the input using an L2 norm and multiplies the output with scale value. The scale can be equal for all channels or one scale per channel. More... | |
| struct | cldnn::permute |
| Permutes data in the memory, with respect to provided order. More... | |
| struct | cldnn::pooling |
| Performs "pooling" operation which is a form of non-linear down-sampling. More... | |
| struct | cldnn::prior_box |
| Generates a set of default bounding boxes with different sizes and aspect ratios. More... | |
| struct | cldnn::proposal |
| struct | cldnn::region_yolo |
| Normalizes results so they sum to 1. More... | |
| struct | cldnn::reorder |
| Changes how data is ordered in memory. Value type is not changed & all information is preserved. More... | |
| struct | cldnn::reorg_yolo |
| Normalizes results so they sum to 1. More... | |
| struct | cldnn::reshape |
| Changes information about inputs's layout effectively creating new memory which share underlaying buffer but is interpreted in a different way (different shape). More... | |
| struct | cldnn::roi_pooling |
| struct | cldnn::scale |
| Performs elementwise product of input and scale_input. More... | |
| struct | cldnn::softmax |
| Normalizes results so they sum to 1. More... | |
| struct | cldnn::split |
| Performs split operation on input. More... | |
| struct | cldnn::upsampling |
| Performs nearest neighbor/bilinear upsampling Also supports built-in Relu activation available by setting it in arguments. More... | |
Enumerations | |
| enum | cldnn::prior_box_code_type : int32_t { corner = cldnn_code_type_corner, center_size = cldnn_code_type_center_size, corner_size = cldnn_code_type_corner_size } |
| Select method for coding the prior-boxes in the detection output layer. | |
| enum | cldnn::eltwise_mode : int32_t { cldnn::eltwise_mode::sum = cldnn_eltwise_sum, cldnn::eltwise_mode::sub = cldnn_eltwise_sub, cldnn::eltwise_mode::max = cldnn_eltwise_max, cldnn::eltwise_mode::prod = cldnn_eltwise_prod } |
| Select mode for the eltwise layer. More... | |
| enum | cldnn::pooling_mode : int32_t { cldnn::pooling_mode::max = cldnn_pooling_max, cldnn::pooling_mode::average = cldnn_pooling_average, cldnn::pooling_mode::average_no_padding = cldnn_pooling_average_no_padding } |
| Select method for the pooling layer. More... | |
| enum | cldnn::upsampling_sample_type : int32_t { cldnn::upsampling_sample_type::nearest = cldnn_upsampling_nearest, cldnn::upsampling_sample_type::bilinear = cldnn_upsampling_bilinear } |
| Sample mode for the upsampling layer. More... | |
Detailed Description
Enumeration Type Documentation
◆ eltwise_mode
|
strong |
Select mode for the eltwise layer.
| Enumerator | |
|---|---|
| sum | Eltwise sum. |
| sub | Eltwise subtract. |
| max | Eltwise max. |
| prod | Eltwise product (Hamarad). |
Definition at line 32 of file eltwise.hpp.
◆ pooling_mode
|
strong |
Select method for the pooling layer.
| Enumerator | |
|---|---|
| max | Maximum-pooling method. |
| average | Average-pooling method - values. |
| average_no_padding | Average-pooling method without values which are outside of the input. |
Definition at line 32 of file pooling.hpp.
◆ upsampling_sample_type
|
strong |
Sample mode for the upsampling layer.
| Enumerator | |
|---|---|
| nearest | upsampling nearest neighbor. |
| bilinear | upsampling bilinear. |
Definition at line 32 of file upsampling.hpp.
