These ones:
https://hybrid-vocal-classifier.readthedocs.io/en/latest/reference/features/named_features.html
https://github.com/vocalpy/hybrid-vocal-classifier/blob/main/src/hvc/features/tachibana.py

Probably we should avoid using someone's name to refer to a feature group ... prefer to use the name from the original MRI paper, I think they actually give a name to their whole set of features IIRC?

... ah yes, CUIDADO. "A large set of audio features for sound description (and classification) in the CUIDADO project". Que nice que le den un nombre asi en espanol.

But to let someone reproduce the paper it would be helpful to refer to the group of features by name like TachibanaEtAl2014.

One thing that will be tricky about the features are defined as scalar/vector values from a segment instead of scalar values from a frame. Maybe we could have a wrapper like extract_scaler_value or something