Nano Edge AI Studio Validation

[xinmeng_wit]

 

I encountered an issue while using the NanoEdge AI Studio tool for Validation.

In step 5, three files need to be uploaded: File Learn, File Regular, and File Abnormal.

What data is Flie Name?

 

 

[Julian E.]

Hello,

 

Anomaly detection projects have a unique feature which is the learning on device (you can retrain the model directly on the board).

What you get from the benchmark is the structure of models that works well for the data used, but not a pre trained model. You need to use "learning signals" to retrain the model.

In the validation step, the knowledge is not available, so you need to use a "learn file" containing nominal data, for the model to retrain itself and then two other files to do the tests.

It is advised to use a learning file containing nominal signals different than the nominal file. You can import the nominal data from the benchmark, but use new files for the "regular file" and "abnormal file".

 

Later, when compiling the model, you have the choice to include, or not, the knowledge from the benchmark:

• If you use the knowledge, it will use it as learning data and work as in the benchmark.
• If you don't, you will need to learn new nominal signals to retrain the model. 
• You can do both if you want. You can complete the knowledge from the benchmark by learning other signals later.

.

A use case example of the learning on device is the following. Let's say that you have multiple machines of the same model. Some are newer than other, so they will have slightly different signals. With the learning on device, you can collect data from few machines, use the benchmark to get the model and deploy the same model on all machines. The trick is that you can now learning nominal signals from each machine and get model specifically trained for each machine!

 

Best regards, 

Julian

[xinmeng_wit]

Hi Julian,

Thank you very much for your answer.

 

I still don't understand the significance of learning on the device.

If there are 100 devices, do I need to upload 100 different files on my PC? This is also a very frustrating operation.

Can I skip to deployment without validation?

[Julian E.]

Hello,

 

Yes you can go to deployment directly after the benchmark.

The validation is made to make sure that the library selected during the benchmark is indeed the best one. To do that we use the validation to test multiple libraries on new data

The emulator is also there for this purpose. It is made to test more robustly a specific library.

 

Here is an example of how learning on device could be interesting.

You have 100 machines, but you don't want to log data on every single one machine. You can log data on let say 2 to 10 machines. Use Nanoedge to find the best library and then deploy the same library on every machines. But in your C code, you will have a part to log few learning signals and retrain the library before doing detection. It allows each library to be more specifically train for each machine.

 

Best regards

Julian