Execution the model with new input image

[sktm502]

Hi.
I followed the instructions and succeeded to 'Putting everything together' on this page.
(https://github.com/google-research/deeplab2/blob/main/g3doc/setup/getting_started.md)

What I want to do is to see the segmentation results for new input images using the result model of running 'export_model.py' (saved_model.pb and so on).
But I don't know how to utilize the results of 'export_model.py'. It would be nice to know the relevant example code.

For information, the dataset what I used is Cityscapes and used pretrained checkpoint about ResNet-50 bakcbone (Cityscapes Panoptic Segmentation) in here
(https://github.com/google-research/deeplab2/blob/main/g3doc/projects/panoptic_deeplab.md)

Thank you very much.