I need to run VGG16 as a convolutional base for feature extraction on some images. I am trying to initialize it on the Raspberry Pi 3B, but the Pi keeps crashing and restarting. I suspect that the Pi doesn't have enough memory for VGG 16, but how should I fix the problem. Here is the VGG16 initialization:
conv_base = VGG16(weights='imagenet', include_top=False, input_shape=(320, 240, 3)) # 3 = number of channels in RGB pictures
I can provide more code if needed.