MultilevelCropAndResize_TRT

MultilevelCropAndResize_TRT - 1

Version

• name: MultilevelCropAndResize_TRT (GitHub)

• domain: main

• since_version: 1

• function:

• support_level:

• shape inference:

This version of the operator has been available since version 1.

Summary

Multilevel Crop and Resize TensorRT Plugin.

Attributes

• image_size (required): Image size. Default value is ?.

• plugin_version (required): Version number of the TRT plugin. Default value is ?.

• pooled_size (required): Pooled size. Default value is ?.

Inputs

• boxes (heterogeneous) - T: The boxes input tensor.

• feature_map_0 (heterogeneous) - T: The first feature map input tensor.

• feature_map_1 (heterogeneous) - T: The second feature map input tensor.

• feature_map_2 (heterogeneous) - T: The third feature map input tensor.

• feature_map_3 (heterogeneous) - T: The fourth feature map input tensor.

Outputs

• patches (heterogeneous) - T: The cropped patches output tensor.

Examples