Hello! I am facing issues on how to trigger events with CLI .eim files. The socket connection does work, however I am having troubles on discouvering how to access the results from my inference. I would like to analise the labels and keep it in a variable. btw, I am using the rasoberry pi 4 with the keyword spotting system model, and therefore, I would like to analise the results to activate a command. It would be really helpful if yave any information!
Hi @catmorais_07,
I’m guessing you are using this doc as a guide? Linux EIM Executable | Edge Impulse Documentation.
What language are you using to create your application? If it’s one of the supported SDK languages (Python, Node.js, Go), I recommend using one of the SDK libraries to perform inference, as it makes using the .eim runner much easier. For example, here is how you can use the .eim file with Python to perform inference: linux-sdk-python/examples/custom/classify.py at master · edgeimpulse/linux-sdk-python · GitHub
Hi! Thank you very much!!!
I have deployed my impulse in a C++ SDK! however I am having some issues… I keep receiving the massage: failed to read audio data. Do you have some idea of the solution?