I’m new to machine learning, but I wanted to try to create my own detector. I know that they often resort to transfer learning, but I would like to try to implement my own model. The problem is that after training (only one class), the model produces coordinates for one supposed object, but there may be several of them. Although such models as yolo, retinanet, etc., find several objects in the image at once.
Does anyone know how this can be implemented?
Please check out this link to see how you can improve your question: stackoverflow.com/help/how-to-ask