Train more object detection models on top of existing model


When I trained a custom model using on top of the prettrained model, SSD mobilenet, why can’t my newly created savedModel be able to detect some of the pretrained objects such as person or car?

I do not know if it is possible to add new object detection model to an existing dataset (inference graph)? Please advice on how I can do that. Thank you

You can add new objects to the dataset but of course you need to train the model again.

Not detecting some objects depends on what you are doing. If you fine tune ssd-mobilenet with flowers only for example, all the other classes that the base model “knows” won’t be recognized anymore as they are not part of the classifier layers anymore (only the flowers are present).