Hello guys, i have applied the finetuning example (TorchVision Object Detection Finetuning Tutorial — PyTorch Tutorials 2.2.0+cu121 documentation) with pedestrians and I applied it with my own object. In the example, the code differs only 1 class of person. But what i need is to train the ground, and 2 different objects (A and B)
These are the different objects:
I have trained the object A and with this model i were able to predict both objects (A and B) but i dont want it!
I would like to differ the models and with modelA i dont want to recognise the object B. I only want to recognise object A with modelA. What should i change, the number of classes?
Thank you very much!