object detection

I am trying to detect tennis balls using the H7 camera. I followed every step on the Edgeimpulse tutorial on how to train my data and I did 9696 for the image width and height however when I want to run my code it gives me an error saying “MemoryError: Out of normal MicroPython Heap Memory! Please reduce the resolution of the image you are running this algorithm on to bypass this issue!” I’m using the exact same code that edge impulse gives you to test your data after you are done with the training. I have tried reducing the resolution but the same error occurs. I repeated the training again but I did 3232 but this time I get an error saying “OSError: Failed to allocate memory. Requested: 5120, available 2816, missing: 2304” Could someone please help me.

Hi, the H7 can’t handle a network that big. Please reduce the network resolution to 48x48 pixels.

Hi, I am facing the same problem when i was modifying the person detection file with my tflite file. Can I know how to reduce the betwork?