Advanced enough to recognise any words it has been trained on.
For example, these offline models can recognise commonly used, conversational words, with a high degree of accuracy(99% in general usage). Because they have been trained on those words. It will get more inaccuracies when trying to recognise unfamiliar scientific/technical words.
Whisper+, FOSS offline voice-recognition.
Pretty advanced! Before it was modern, it wasn’t quite as advanced
It’s pretty damned good for generalist things. Especially after training on your voice. I’m quite fond of gaming with Voice Attack - for those unfamiliar think saying “red alert” or “evasive action” and a bunch of macros instantly run a pre-programmed sequence of keyboard, mouse and joystick commands.
It still fails for the higher vocabulary of well educated professionals and specialist lingo. It also chokes on accents. and this.
Better than it was in 1997?
Fudo voice works very well for English language on my Android. I’m using the largest language model.
did you mean futo voice? if so, agree with you. I’m constantly blown away by it
