How can I download baby crying sound dataset on LPC54114.

[Robert9]

Hello.

How can I get LPC54114 learned baby crying sound?

or How can I download baby crying sound dataset on LPC54114.

Where can I find reference page?

Thank you very much.

[Robert9]

I will use microphone for recognition of my speech that is recorded to th kit.

[frank_m]

> ...baby crying sound ...

Redundant information. You basically mean "a specific sound".

Either record it, or find it anywhere on the internet. I think even Google didn't censor crying babies yet.

Convert it to a file format you want to support, and add it to the project ressources. I think there are even tools to create C structures/arrays from such data.

> How can I get LPC54114 learned baby crying sound?

This is less trivial, pattern recognition in audio data. Speech recognition is closely related. Perhaps you can use spectral tools (FFT). There are quite a few articles on the net, but no free library I'm aware of.

Most commercial baby monitor devices operate on intensity (volume) alone. At least they did when my kids were at that age ...

[Robert9]

Thank you for your kind answer.

Can I ask another question?

According to SDK of LPC54114 Audio and Voice recognition Kit,

maybe I can record my speech to the device.

If so, how can I  get the kit recognized my speech?

Thank you very much.

[frank_m]

I don't know this specific kit, especially the SDK. I would suggest to read through some SDK documentation first, and try some examples first.

> maybe I can record my speech to the device.

I am pretty sure this is the very purpose of the kit. So this should work. I am having an OM2098, featuring a LPC54628. Performance should be sufficient for quite a lot of elaborate methods.

> ...how can I  get the kit recognized my speech?

This depends on your interpretation of "recognizing my speech".

Audio and speach recognition is usually about "transforming" audio data to text, i.e. recognizing spoken words - regardless of the speaker.

If you want it to recognize you voice specifically, things getting more difficult. You would need to distinguish your specific intonation and voice harmonics somehow algorithmically.

I think there are some generic resources out in the internet, about the mathematic basics and PC-based libs and examples.