how to load faceNet to get inference from realtime_testing.py ?

[JavanehBahrami]

Hi how to load faceNet to get inference from realtime_testing.py ? I used this way: I load the checkpoint inside the pretrained model that you have added on the net, butit gives me an error

model = load_model("pre-trained_weights/model_inception_facial_keypoints.h5",custom_objects= 
                                  {"Adamw":tfa.optimizers.AdamW},compile=False) # updated the loading function

faceNet = model.load_weights("pre-trained_weights/model_face_net_files-20231223T152222Z-001/model_face_net_files")

resultImg,faceBoxes = highlightFace(faceNet, frame)

the error is:

resultImg,faceBoxes=highlightFace(faceNet, frame)
  File "realtime_testing.py", line 19, in highlightFace
    net.setInput(blob)
AttributeError: 'Functional' object has no attribute 'setInput'

Please, I will be appreciate if you can help me @dsabarinathan

[ahsanMuh]

Any update on this?

[dsabarinathan]

@ahsanMuh @JavanehBahrami , you can load the tensorflow model directy using the below script.

model = load_model("pre-trained_weights/model_face_net_files-20231223T152222Z-001/model_face_net_files",custom_objects= 
                                  {"Adamw":tfa.optimizers.AdamW}) # updated the loading function

print("model input: ",model.input)
print("model output: ",model.output)

[dsabarinathan]

@ahsanMuh @JavanehBahrami I updated the real time testing code with new changes. Please use it for testing.