Hello everyone,
I am detecting different pathologies from a chest x-ray. I trained and evaluated the model. I have model wights the model is denseNet-121.
I am preparing an inference file for the model but the model need is not easy and simple as people suggested in past topics.
I have already check these link
Link 1
Link 2
Link 3
Link 4 etc
Now let’s get into my problem with code explanation.
I have full detection model which have multi-models in that model like densNet, Multi headFPN, and simple CNN and ResNet too. each model has a specific reason and my model has following three main functions like this
models.get_detection_model, models.get_classifier, models.get_detection_learner
but I will call the get_detection_model.
- Now in inference the model needs Databunch and I don’t know how to make Databunch for a single image.
- is I am doing correct as below code?
learn = rps.models.get_detection_learner(data, models.densenet121, crop_arch=models.resnet34, path=path)
learn.load(’/home/ali/Desktop/chest_project/RpSalWeaklyDet/models/fullmodel_densenet121’);
here data is a databunch but I don’t know how to do it?
I clone the code from GitHub repo link is here: Link
What should I do for this please help me fastai.
Note: I am not expert in fastai but worked in keras and tensorflow.
Thanks you