Custom Model Type Detector

[AndrewBrenev]

Hi, I've trained my model and now I'd like to detect this objects. Here I found that "The models supported are RetinaNet, YOLOv3 and TinyYOLOv3".
Is there any custom model type detector, exept this three? Or is there any other way to detect objects on the model I trained? Here ids the code I use:

from imageai.Detection import ObjectDetection
import os
execution_path = os.getcwd()
detector = ObjectDetection()
detector.setModelTypeAsRetinaNet()
detector.setModelPath( os.path.join(execution_path , "model_ex-020_acc-0.500000.h5"))
detector.loadModel()
detections=detector.detectObjectsFromImage(input_image=os.path.join(execution_path,"icon.jpg"),
output_image_path=os.path.join(execution_path , "iconnew.jpg"))
for eachObject in detections:
print(eachObject["name"] , " : " , eachObject["percentage_probability"] )

Compilation of this code gives an error about disparity layers. Thank you for your reply.

[JStuve]

I too am waiting for this. It seems that it was mentioned that this feature will be release in the next update. That was back on Feb 22nd of this year.

[bread-on-toast]

Hopefully there will be progress soon...

[OlafenwaMoses]

Keep up to date with this thread. #8