Splitting audio at word boundary

[smart...@gmail.com]

Dear Friends, Hello.

I am trying to split audio at the word boundary,  for example, audio after each word.

Is this possible with librosa? while I cannot make it with any other python library like pydub.

Like

He is a boy. He works hard at school.

Here while splitting with silence, it breaks mostlikely at sentence. while i want at word.

split at silence is the data-loss as well.

Your kind help is requested.

Greetings.

[Brian McFee]

I wouldn't recommend using librosa (alone) for this.  Segmenting an audio stream of natural speech into individual words requires a high-level understanding of the langauge involved.  (Try doing it yourself on a language you don't speak -- it's difficult!)

I think your best bet here would be to use a pre-trained automatic speech recognition engine.