Making an ai

[Mattyhawk]

I want to code an ai that can recognise noises like a dog barking and put it in a website.
I'm not sure how to do this though.
I guess that I will have to code it in html.
Does anyone have any ideas.

[B.Goode]

I want to code an ai that can recognise noises like a dog barking and put it in a website.... I guess that I will have to code it in html.

The name HTML is a contraction (abbreviation, or acronym, or something like that) of the full title - HyperText Markup Language.

Its purpose is to describe the elements of documents such as web pages.

So the final presentation of the result of your detection of a dog barking might use HTML to generate a modified web page.

But it probably won't be the means by which you process an audio feed and determine that it represents a dog.

Your project sounds similar to the way that 'agents' like Alexa and Siri recognise their own names as a trigger for processing further commands. That 'AI' part seems to be much more challenging than the presentation of a web page.

[Andyroo]

Have a look at http://www.orelia.fr/en/technology/audi ... level.html

The video identifies a drone so it can be done but no idea of the cost or work involved Also read the difference between identification and recognition.

Pattern matching may well be possible on the Pi.

[Gavinmc42]

On a website for external viewing, means internet access so online recognition software?
Lassie using Siri to order dog food for drone delivery?

There are now a few Hats and FPGA boards with microphones.
Tensorflow Lite or something like that might work.

Audio spectrograms are used for bird recog.
FFT's can be used but there is some research showing LPC generates less noise(data) making it easier for the AI.

Start with Speex on Pi's?

[Heater]

I want to code an ai that can recognise noises like a dog barking and put it in a website.
I'm not sure how to do this though.
I guess that I will have to code it in html.
Does anyone have any ideas.

Yes.

HTML is not a programming language. It's a text mark up language. So you ain't writing any AI codes in that.

Get yourself one of these: https://www.seeedstudio.com/Grove-AI-HA ... -4026.html

seedstudio have microphones, cameras and all kind of things to plug into that. It's very cheap. If you cannot wait for that particular Pi HAT board they have other similar but non-HAT format boards.

There is hardware acceleration on there for the AI things you will need like neural nets, convolution and FFT. As faras I understand they have AI libraries that allow you to use it all from Python.

You can always display your results, formatted with HTML, in a web page from a server on your Pi.

When you have figured out how to do it and have a working sound recognition system, please do write about it somewhere and let us know. I'd be interested in getting started with that but don't have the time for it.