BME688 classification python question

[IBS123]

Hi,

I've trained an alogrithm using the BME AI studio and the dev-kit board. Following that the alsogirhtm was exported as a package of files: .csv, .h, .c, .config and .aiconfig

I would like to use the trained algorithm on another BME688 board connected to raspverry pi. 

How can I use the trained algorithm to output the same classes as used in BME AI studio?

Thanks in advance

[BSTRobin]

Hi IBS123,

If you have downloaed latest BSEC2.4.0.0 package, you can refer to the Arduino example code "BSEC2.4.0.0_Generic_Release_23012023\examples\BSEC_Integration_Examples\examples\bsec_iot_example\bsec_iot_example.ino", and replace 'BSEC2.4.0.0_Generic_Release_23012023\examples\BSEC_Integration_Examples\src\config\*.c', 'BSEC2.4.0.0_Generic_Release_23012023\examples\BSEC_Integration_Examples\src\config\*.h' with your configuration files in the example code.

[IBS123]

Hi BSTRobin,

Thanks for the quick reply, I wanted to ask in addition:

1) Could you please provide a link to the example and package to be downloaded?

2) I've looked at the .c and .h files - these are the coefficients of the neural net. The classes are stored in the .aiconfig - how is it used?

3) Is there a ref for python code to be used with Raspberry pi instead of Arduino board?

Thanks

IBS123

[IBS123]

Hi, 

I found the link to download the *.ino file but didn't understand how to use the classification of the trained algorithm and output "class A" or "class B" etc, because the .c file only comprises the coefficients...

Thanks

[kmca · Accepted solution]

I have an article  and  python code on this on github here:  https://github.com/mcalisterkm/teach-your-pi-to-sniff-with-bme688. 

PI3G have a PI/BME 688 video (Cheese  Meat Detector) on you tube that is entertaining.