UNOMI is a SaaS for animators and video game developers. UNOMI leverages advanced, voice recognition and motion capture technology that automates a lot of the most time-consuming aspects of animation production.
I'm sure matching transcription text to audio is a hard problem to solve but it seems like the quality of the results is highly dependent on the accuracy of the _large_ number of face poses provided by the user. In other words, your tool could perform perfectly but if the user-provided face poses are bad, the results will be invariably bad.
Suggestion: It would be better if you provided ready-made 2D/3D face poses that the users could use as the basis for their animation/models. That would not only save users a great deal of work but help increase the likelihood that their results will show off the quality of your tool.
@stephen_jones1 Hello Stephan. Thank you for the response. Yes, our Avatar Creation tool will come with pre-made poses for the characters that they create. Most of our users want to create their own mouth poses to achieve specific looks for their characters.