I have STEVAL-MKSBOX1V1 . I have downloaded the functional pack where shall I find the baby cry detector... Any user manual to understand the AI details and coding details

[Eleon BORLINI · Accepted solution]

Hi @Snara.2​ ,

the AI model and the library source code are not public, so I cannot disclose details of the Neural Network (NN) used for the acoustical recognition of the baby crying pattern. I can tell you that it is based on the extraction of representative features from an audio acquisition (such as the FFT, the mel coefficients, sound intensity, etc...), but not through the "visual" analysis of the spectrogram.

By the way, I can anticipate you that we will publish an article on this topic in the following weeks, that will better explain the audio NN behind the application and the use of the Sensortile.box.

I'll keep you updated about the progression of this knowledge article, for now here you can find some more details.

-Eleon

[Snara.21]

@Eleon BORLINI​ 

Thanks for your feedback. Is there is any business model where we shall sign an NDA with ST team. If so please let me know the right person I need to approach to have business meetings.

[Eleon BORLINI]

Hi @Snara.2​ ,

for this purpose, please contact a sales and representatives office of your geographical region from this LINK.

They will evaluate the case and the opportunity and give you a feedback.

-Eleon