Hacker News Reader

> or notice something weird

> [4]: Sample reconstructions

The text in the reconstructed image in the third row looks different, the word phonomat is quite garbled, information looks a bit funny.

View on HN

Yeah, high frequency detail such as facial features for faraway figures or text tend to get washed out after compression - this is probably due to a couple reasons: 1) The training dataset contains relatively few pictures including text, 2) high-frequency detail is too expensive to encode and the model learns to forgo encoding this in favor of more 'important' features such as shapes, colors, etc.

Reply on news.ycombinator.com