Custom Classifier on top of all existing species in BirdNET

[safahafiz]

Can I build a classifier on top of the existing one for all the species using new data. In other words, I want to add new species to the list that BirdNET already has. However, the custom classifier only tests the input with the species in the custom classifier.

Is it possible to do transfer learning to train BirdNET?

[Josef-Haupt]

Hey, you can indeed extend BirdNET with your own classes. If you're using the GUI check the "append" option under Model Save Mode. In the CLI the option is called --model_save_mode in train.py.
This mode outputs the BirdNET classification and concats them to your custom classifier, so you get one long vector, with your new classes at the end.
Does that answer your question?
text copied from the issues

[FlorianH27]

Can "append" only be used to add new species to the existing model, or can it also be used to improve the performance for already implemented species?

[Josef-Haupt]

The current implementation is "stupid". It let's you train whatever classes for your new classifier and just concats the resulting vector to the BirdNET output. So redundant classes are possible, you just have to use the correct one in the vector.

[w-out]

A related question on using the "append" option. Whenever I try to run my appended model/classifier on multiple files, I get an error in the GUI. The only way it will work is when I manually remove the double species names from the species list that is created during the training process. However, I am afraid that removing these double species names will influence my results, as I get approximately the same results with my appended classifier compared to BirdNET's own model.