Sending NN inference via WiFi
Posted: Tue Jul 13, 2021 2:16 pm
Hello!
I'm working on a project where I will be deploying a NN model (image classification) to the AI-Deck. I want to send the classification results back to a PC through NINA's WiFi communication. Is there a possible way to do that by changing the JPEG Streamer example code scripts (viewer.py, ai-deck-jpeg-streamer-demo.c, test.c)?
In ai-deck-jpeg-streamer-demo.c, you can see that GAP8 sends a special request packet to the NINA module (type 0x81). Do you have an idea how does GAP8 formulate those NINA requests, where and whether they can be manipulated/changed to perform my desired task?
I'm working on a project where I will be deploying a NN model (image classification) to the AI-Deck. I want to send the classification results back to a PC through NINA's WiFi communication. Is there a possible way to do that by changing the JPEG Streamer example code scripts (viewer.py, ai-deck-jpeg-streamer-demo.c, test.c)?
In ai-deck-jpeg-streamer-demo.c, you can see that GAP8 sends a special request packet to the NINA module (type 0x81). Do you have an idea how does GAP8 formulate those NINA requests, where and whether they can be manipulated/changed to perform my desired task?