Re:Image Scaling Using Deep Convolutional Neural Networks
by ska on Wednesday May 06, 2015 @09:15AM
(#9498325)
I wonder how this approach compare to upscaling via IFS (i.e. "fractal compression"-like techniques). Although the underlying modelling is very different, the fundamental idea seems very similar - learn vectors from the image, apply (possibly a mixture of) them at a different scale. The results certainly look superficially similar.
[ Parent
| on HN ]