Live data from Hacker News

Nearly 1,000 Paintings and Drawings by Vincent van Gogh Digitized and Put Online

openculture.com

81–90 of 93 posts

Re: Nearly 1,000 Paintings and Drawings by Vincent van Gogh Digitized and Put Online

#81
post #52

Earlier quoted context omitted.

I really don't understand your comment, and all subsequent responses along the same line. The digitized van Gogh paintings are perfect copies of the original public domain works. For the US, Bridgeman Art Library v. Corel Corp. [1] makes it perfectly clear that the images are not protected by copyright, and there is similar law and legislation in Europe. What do you mean with the ability of the museums to choose a li…

I believe the problem lies in the wording of ‘exact’. Without a direct comparison to the source material ‘exact’ could be challenged. There’s also the problem of exact dating of a piece of art as being pre-1923. With a reproduction in a published book you have a publishing date in the colophon that logically must be the same year or later than the reproduction of the artwork inside. Without that you’re relying on the…

What if I download an "A4" and distribute an "A3" with exactly the same pixels?

Re: Nearly 1,000 Paintings and Drawings by Vincent van Gogh Digitized and Put Online

#82
post #81

Earlier quoted context omitted.

I believe the problem lies in the wording of ‘exact’. Without a direct comparison to the source material ‘exact’ could be challenged. There’s also the problem of exact dating of a piece of art as being pre-1923. With a reproduction in a published book you have a publishing date in the colophon that logically must be the same year or later than the reproduction of the artwork inside. Without that you’re relying on the…

What if I download an "A4" and distribute an "A3" with exactly the same pixels?

No idea, and that’s the problem: the licence is badly defined. Ideally they would have used one of the pre-existing CC licences that people already understand.

Re: Nearly 1,000 Paintings and Drawings by Vincent van Gogh Digitized and Put Online

#87

Script to download full sized images: https://gist.github.com/RyanEager/a747215c259ffecb9beb1dfec0...

What kind of dimensions do you get from the tiled downloads? I grabbed a few of the 'large' JPGs the site offers and wonder how much of a difference there is.

Re: Nearly 1,000 Paintings and Drawings by Vincent van Gogh Digitized and Put Online

#88

Earlier quoted context omitted.

> There's no way to transfer consciousness from the biological to the electronic substrate Transferring? You are thinking small. What about expanding beyond your brain, so that it becomes a small part of a whole and by the time it starts failing it's just a peripheral input preprocessing device/local motion planner/unreliable redundant memory storage/small vote in global decision making and not you. The possibility i…

How about slowly replacing brain cells with synthetic ones while maintaining the consciousness in place?

Technical difficulties will be much greater. The brain is still a single point of failure. The "expansion" approach allows gradual increase of intelligence thru intermediate exocortex stage, boosting the progress, while studying which information processes/something else corresponds to conscious subjective experiences, paving the way for relocation of consciousness locus outside of the brain or resorting to other measures in a case of impossibility.

Re: Nearly 1,000 Paintings and Drawings by Vincent van Gogh Digitized and Put Online

#89

Slightly off-topic, a big part of seeing a painting in real life is seeing the structure and layers of brush strokes. It's a long shot but does anyone know if any work has been done to create depth maps of paintings and possibly combine them with scans?

https://verusart.com. 3D printed. $1,400.— buys you a copy of a van Gogh (https://shop.verusart.com/collections/verus-art-national-gal...)

Re: Nearly 1,000 Paintings and Drawings by Vincent van Gogh Digitized and Put Online

#90
post #6

Woah. In an ML class I took last fall we got to design our own classification challenge problems that the entire class then competed on. The one our group made used this exact dataset (the museum website)! We called it VANGOGHORNO haha. We basically pulled the images from the collection and mixed them with non-Van Gogh paintings. It was surprisingly a very learnable dataset, the top two teams got around 97% accuracy…

”It was surprisingly a very learnable dataset”

It also is a dataset where cheating is quite easy. A google search may quickly turn up items you left out of the training set.

Do you know whether any teams did this (could even have happened by accident, if people googled to find different views of paintings, and accidentally included pictures not in the training set)?

Post reply on HN