@krmneimneh :D a indeed. In concept in terms of usercase it is similar to Shazam but mathematically what we are attempting to do is much more advanced.
Shazam can't recognize original audio/sound...you can't sing a cover of Adel's Hello waiting for it to recognize the song, that's because it uses sampling...fingerprint matching the audio file in its insanely large database.
Also we will be working with medical professionals this year after closing our seed round to conduct clinical trials, the largest of its kind every tied to prelinguistic communication....the response for this has already been extremely positive.
With Luna the ambitious nature was to use AI and deep learning to intelligently analyze the cries, sounds of infants even if we haven't fully sampled it and for the system to come up with an answer on it's own learning as it grows/in time.
The science behind this is called infant cry analysis or pre linguistic communication it's been around since the early 20s.
When it was really established academically...computers personal computers weren't even invented....and the computing power was laughable by our standards now, that's the transformative element we can put this in the hands of parents everywhere.
Our mission is to empower and strengthen our connection toward each other by fostering delightful experiences around technology.
This couldn't be more true to the tune of the mother and child relationship...and starting to have a completely new conversation around communication.
Because infants can communicate we just until now haven't been able to put something in the hands of parents to interface with...we have done so academically, lots of research on this subject.
I'm intrigued since I have a 10 month old daughter and I'm still figuring out this parenting thing! I am curious to know if this app only applies to newborns? I wasn't able to find any info on the age range this app supports.
@abadchris Hey Chris thanks for pointing that out, I hear you man and that's what luna is all about as a brand trying to empower you to do just that...we're going to update the description. it works for infants so any child under 12 months but the effectiveness/accuracy at least for now varies. The younger your child is the more accurate it is as of right now. Raising a funding round will change this ;)
Issue right now is our system has to learn not just about the classification system but to tailor the deep learning to your individual child that's the big transformation which is why we had to launch this in an MVP state to start training it a bit more.
For an infant around 10+ months accuracy we're looking at accuracy around 70%
All of the research has gotten this to 85% and 93% respectively and that was 50 years ago, and for disease so we're confident we will be able to this year get it to around 90-95% for up to 12 months..that's our goal.
Got A Minute?
Castle
Castle
Castle