My coreml model has NWHC input shape.(Number, Width, Height, Channel) I want use it on Vision with Image input. But when I change its input from multi array type to image ty
