How to use my own pretrained model to detect custom object after training

Well, i trained an SSD detector following the example

https://www.mathworks.com/help/vision/ug/object-detection-using-single-shot-detector.html

then i saved the detector using save(detector) and it got saved under the name 'detector.mat'

but when i try to use it again after shuting my laptop,

detector = load('detector.mat')

matlab doesn't recognize and it tells me it's either undefined or parameter wrong type. however i can use the matlab pretrained network that i downloaded

Please how can i use my own pretrained network?

Answer 1

[Kei Otsuka]

save('detector') saves all variables from the current workspace. So, you need to extract detector from all variables after loading mat file like this;

vers = load('detector');
vers.detector

Or, please try to use following syntax to save only the detector instead.

save('detector.mat', 'detector')