Chaining together processing blocks

#1

Hello!

I am doing some work on an IMU gesture detection project.

I would like to do some pre-processing on my data before I pass it over to the spectral feature processing module. It is not clear to me whether chaining processing like this is supported? If so, how do I configure it?

Thanks!

Ben

#2

Hi @Benjii519

What tools are you trying to integrate with or are you chaining multiple of our Learn / Processing blocks? Yes both is possible, if you are coming from an ML background try using our Python SDK:

We also have other apis available and more advanced offerings in our enterprise tier if this is for use in production.

Best

Eoin

#3

Thanks!

I was looking to do some pre-processing followed by the “Spectral Features” module in the web interface. I found the instructions for creating custom processing blocks, and was hoping I could then route the output of that through the web interface’s “Spectral Features”

So looks like the answer is “not from the web interface” but absolutely through this SDK.

In the end, I just ended up adding the pre-processing on device, since it is relatively simple. But this is useful to know, esp. if I want to change up processing on data I’ve already collected.

1 Like
#4

Great!

Please share that with us if you think it can be useful for others or something we can add to the firmware.

Also it you want to do an enterprise trial add a company email and we can get one set up so you can explore our other preprocessing options with custom blocks.

Best

Eoin