Hi!
I want to perform semantic segmentation of a series of input image using a deep learning model.
I have an already trained CNN model in MATLAB. The format of that trained model is in ".mat". I want to load that trained model into LabVIEW and need to perform semantic segmentation for a given input image(similar to images used for training the model). I can find examples explaining how to load a frozen graph of a trained model from a file in .pb format. But not getting any sources related to loading a trained model in .mat format.
Then I want to perform semantic segmentation using that model. I can see there are some advanced VIs in Vision and Motion-->Machine VIsion-->Machine Learning-->Deep Learning. There are vis only for image classification and object detection. How to perform semantic segmentation in this case.
Questions:
1. How to load a saved trained Deep learning model in '.mat' format into LabVIEW?
2. If succeded in this loading of the trained model, which vis need to be used or what framework need to be used, in order to perform semantic segmentation.
Any type of suggestions is appreciated.
Best
Arya
