Swift Tutorial: Native Machine Learning and Machine Vision in iOS 11
Alex Wulff

Without having seen the code, i’m guessing the network was trained, using images of a fixed size. The network may prefer smaller image sizes, because it was trained using smaller images. The network may be too specialized, also known as “overfitting”, which in this case, means it doesn’t generalize well to other image sizes.

But that’s all guesswork. No facts here ;)

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.