I love cats, and over the years have noticed that they have recurrent patterns of vocalizations. For example, upon seeing a bird, a cat may start chittering, but the same cat would never chitter at humans. Then there are complex vocalizations, like meow-wow, which I have observed across multiple cats on different continents. At the same time, we have birds and monkeys which have vocabularies of up to 300 (?) words. It seems like cats are communicating something, but humans may be too tone-deaf to understand that.

It seems to me like the task of understanding what a cat is trying to communicate to humans is suitable for some kind of machine learning process.

My question is: has any of these unsupervised models been applied to cat vocalizations? In other words, if a model can draw or generate text, can it generate cat meows? How close are we to understanding what cats are meowing about and translating it into English?

I remember that some work has been done with trying to decode dolphin vocalizations, but as you can imagine, that requires specialized equipment, while a cat model can be tested in the real world with simpler equipment.



You must log in to answer this question.