Skip to content

Latest commit

 

History

History
 
 

pix2pixHD

pix2pixHD

Input

Notice: The current streetview model need both label map and instance map as input, since the original pytorch model was trained with both maps as input.

Input label map

Input instance map

Standard shape of 2 maps: (1, 1024, 2048, 1) grayscale, in the format B, H, W, C

If input shape is not matched, it will be reshaped to : (1, 1024, 2048, 1)

(Images from https://www.cityscapes-dataset.com/)

Output

Usage

Automatically downloads the onnx and prototxt files on the first run. It is necessary to be connected to the Internet while downloading.

For the sample image,

$ python3 pix2pixhd.py

If you want to specify the input label map and instance map, put the label image path and instance image map after the --input and --instance_map options respectively. You can use --savepath option to change the name of the output file to save.

$ python3 pix2pixhd.py --input IMAGE_PATH --instance_map INSTANCE_PATH --savepath SAVE_IMAGE_PATH

Reference

Framework

PyTorch

Model Format

ONNX opset=11

Netron

pix2pixhd.onnx.prototxt