I am self learning developer so forgive me if I am asking something obvious or not clear with my request. I was looking forward to create an app and in that I want the user to speak a word and that would trigger an event.
I think sensitivity of recognition is not essential as the app only needs to check if the word was spoken or not, rhyming words will also do but it needs to be fast and better offline.
For online voice recognize, there are lots in the internet. you can find several in asset store (Unity Asset Store - The Best Assets for Game Making). It’s usually just send the audio to OS and the device OS usually send it to their company’s server to do recognition. Although, Windows may do it on-device. I haven’t tried.
For offline voice recognize. so far i can only find Keen Research (https://keenresearch.com/ ). I tried it, it’s quite good (you need to preset some words) but it’s quite expensive.
A quick follow up to the previous comment about KeenASR SDK for on-device speech recognition for mobile devices – I am the CEO of Keen Research.
While we don’t publicly discuss pricing information, I can tell you that the licensing model is typically a yearly licensing fee per product (app in this case) which covers all of the platforms we currently support (iOS, Android, soon Linux), access to online development tools, SDK updates, etc. The exact pricing depends on a number of factors.
Comparing to various assets in the Asset Store, you can probably consider our SDK to be expensive. But, it’s all relative to what you are building, who you are, etc. We have a number of customers that are using the SDK in their products.
If you are working on a cool project/game, don’t need a lot of hand-holding, and are interested in using the SDK, please don’t hesitate to get in touch. And, if your project is for good cause (non-profit, etc.), pricing can be very inexpensive.