Hi! I'm an undergrad student and I plan to develop an app for my thesis that has a speech recognition feature with game elements. The app is suitably for children where they can enhance their speech capabilities. I plan to use pocketsphinx as my speech recognition api since I've read it can be used offline and I would like to ask some questions first to secure if some things are possible during the development...
Can I use pocketsphinx within the Unity engine on a Windows OS?
(im using Windows and I plan to develop the app only for the android platform-ive searched for demos for this and so far I havent seen any that's working, only for OS X. And I've also searched within forums relating to the problem and from what I can gather, many problems/difficulties have been come across when using with Unity )
The game features of the app is that it would have basic speech exercises like saying words shown in the screen which is categorized into types (e.g colors, shapes, animals,etc..) and the user would say those words back. I plan to implement a scoring system based on how many times the child got the correct answers...Another type of exercise I plan to include is to make simple conversations with the child user (Greetings, Introductions to get an idea), and it also has a storytelling exercise where the app will get to ask questions about the story (Who're the characters in the story, etc...). Is that kind of feature possible using pocketsphinx? If so, do I have to build a language model or only a dictionary will suffice?
I'm sorry if my questions are kind of stupid and there might been already answers for these types of questions..I'm pretty much a novice with speech recognition and CMUSphinx in general so I hope this discussion forum can help me! And perhaps suggest an alternative speech recognition API for which I can use is also helpful..Have a nice day :)
Last edit: Kim Jones 2018-07-27
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi! I'm an undergrad student and I plan to develop an app for my thesis that has a speech recognition feature with game elements. The app is suitably for children where they can enhance their speech capabilities. I plan to use pocketsphinx as my speech recognition api since I've read it can be used offline and I would like to ask some questions first to secure if some things are possible during the development...
Can I use pocketsphinx within the Unity engine on a Windows OS?
(im using Windows and I plan to develop the app only for the android platform-ive searched for demos for this and so far I havent seen any that's working, only for OS X. And I've also searched within forums relating to the problem and from what I can gather, many problems/difficulties have been come across when using with Unity )
The game features of the app is that it would have basic speech exercises like saying words shown in the screen which is categorized into types (e.g colors, shapes, animals,etc..) and the user would say those words back. I plan to implement a scoring system based on how many times the child got the correct answers...Another type of exercise I plan to include is to make simple conversations with the child user (Greetings, Introductions to get an idea), and it also has a storytelling exercise where the app will get to ask questions about the story (Who're the characters in the story, etc...). Is that kind of feature possible using pocketsphinx? If so, do I have to build a language model or only a dictionary will suffice?
I'm sorry if my questions are kind of stupid and there might been already answers for these types of questions..I'm pretty much a novice with speech recognition and CMUSphinx in general so I hope this discussion forum can help me! And perhaps suggest an alternative speech recognition API for which I can use is also helpful..Have a nice day :)
Last edit: Kim Jones 2018-07-27