I built an app that relies on recognizing one phrase and then performs a certain action if it hears that phrase. I’m having some troubles with recognition from a distance though or in loud-ish environments. Please note that the users will be using my app from a distance most of the time.

I would expect this behavior from a framework that tries to get the transcription from the entire set of audio samples, but I was hoping for something more specific where it would ignore all sounds except for my one phrase. Is this functionality possible or is there something that’s close to the behavior I’m describing?

Thanks,
Derek