Azura’s voice recognition update
It’s been a long time I didn’t do an real big update and there are a fair amount of new stuffs.
I had a lot of fun with SSP voice recognition feature. I found it pretty useful to be able to change a song and to control the music player without having to touch or to be near the computer. So I implemented a menu for the voice recognition. It’s a simplified menu with short commands. Only the settings part isn’t simplified yet.
There are three modes too :
- the full mode : Azura would react to every command you said such as her name, the music player or any choice that is currently displayed and selectable.
- the disabled mode : she won’t react to anything.
- the partial mode : only the music player would react.
The purpose of this modes was allowing people to sing loudly in front of the computer without having Azura reacting at all. People who use voice recognition know what I mean (it’s quite annoying when it reacts for nothing because it didn’t understand correctly what has been said).
Since some song name can be too much long and complicated to be said at once with the voice recognition, I added a renaming feature. So people can rename their songs. It won’t rename their file, only the way Azura call them in her playlist. This nicknames are saved when the playlist is and are restored when it’s loaded.
The AI now, I progressed with the self-learning feature. I implemented the responses for the self-learned words for the normal sentences. So it still misses the questions but she shouldn’t say an error message for the normal sentences.
Still for the AI, when a sentence wasn’t having a verb such as simple greeting, Azura was saying an error message. She shouldn’t anymore.
And the last new stuffs added are new dialogues. I completed the dialogues I started to implement last update.
