Is it possible to perform N-classification using the ISM330IS? If so, how can it be done?

[Federica Bossi]

Hi @Semolina,

Welcome to ST Community!

If the question is related to NanoEdge.ai, no, you can only do anomaly detection.

In general it is programmable, so you can make yourself a decision tree classifier with n classes.

If this helps you, please mark my answer as "Best Answer" by clicking on the "Accept as Solution" button, this can be helpful for Community users to find this solution faster.

[Semolina]

Thank you, @Federica Boss. I possess a TensorFlow Lite (TFLite) model. Could you guide me on how to deploy it on the ISPU? I noticed in the whitepaper discussing ISPU, they accomplished tasks like digit gesture recognition and fan anomaly detection using QKeras. Could you provide insights into how they executed and interpreted the model on the ISPU platform?"

[Amith_lal]

I'm also trying to do the same ! Did you try implementing tensorflow ?

[Semolina]

Yes, but it was to big. I used sciikit learn models ( decision tree and naive bayes) and implemented the classification functions on the ispu by myself.

[Amith_lal]

ok

[Amith_lal]

What are the steps involved in this?

[Amith_lal]

@Federica Bossi Can you provide any document or steps involved in the deploying of ML in ISPU