256 Shades of Gray [pdf] | Hacker News Reader

256 Shades of Gray [pdf]

256shadesofgray.com

34 pointssokrates37 comments

Open articleView on HN

I love the arsclassica latex style!

http://www.ctan.org/tex-archive/macros/latex/contrib/arsclas...

Shhhhhh! Remember the comment about Riff-Raff :)

???

hsmyers is referring to the first sentence in the linked pdf.

    Parallel machines are hard to program, and we should
    make them even harder – to keep the riff-raff off them.
    — Gary Montry

Thanks mertd. Can you explain what this has to do with my latex comment? I was not supposed to describe the magic behind the typesetting/layout?

I believe hsmyers was jokingly trying to say that it would suck if everybody ("riff-raff") started using that template going forward, rendering it "normal", hence not special anymore.

Thanks! I had just come here to ask about the typesetting on this... it's pretty lush. Good job.

The 256shadesofgray.com website is currently work in progress. There will be more details and stuff there soon.

Video of the 256 Shades of Gray machine in action: http://www.youtube.com/watch?v=pOXzD_7pL4M&list=PLA329EE...

An enjoyable paper, a thoughtful example of engineering under constraints, and a successful outcome. Congratulations!

(But, and I'll gladly take the karma hit for saying this, please don't pay any attention to my fellow HNers complaining about every possible tangential non-issue.)

Thats cool, but it seems to have the performance equivalent of an early P4. Theres talk about FPS, but then its mostly about static images?! Theres little information about the actual performance of the system on the FPGA because everything is seemingly hamstrung by the woefully inadequate AVR used for feeding the data.

Sheesh. It's an undergrad term project. They're not trying to line up an IPO.

We didn't really do any benchmarks that would permit me to actually say it has the performance equivalent of anything, but I was probably not alive when the last "equivalent" system was made. This was purely an educational endeavor. There was much talk about the FPS because the class had to separate groups with the same assignment, one was to focus on energy efficiency, and the other (us) on performance.

And yes, the AVR proved woefully inadequate as you point out, but we were required to use it so not much we could do there. The project was loads of fun (and work!) anyways, and it was great to be able to put theory into practice for once :) If anyone have any questions regarding the project I'd be happy to answer them by the way.

This submission needs a better title.

Yes, the name of this project is "256 Shades of Gray" – with an 'a' instead of an 'e' in Gray.

To the implementers: congratulations on a successful project, but please make sure you consider the ethics of your field before applying what you have learned for money.

Am I the only one that is dissatisfied with all of the Lena references in the computer vision community? For a few years I was righteously indignant about the gender issues involved but now it's just trite. Makes us seem like adolescents that we find it so important to reference a Playboy photograph.

rolls eyes

It's just a standard image of a woman wearing a hat that has a lot of dynamic colors and allows people to check and compare graphic transforms easily - there is nothing offensive about it. Unless, of course, you find colorful pictures of women wearing hats offensive.

Erm, you might want to take a look at this (NSFW):

http://www.lenna.org/full/len_full.html

Except that's not the complete story, is it? The complete story is that Lenna is from a culture where nobody makes that big a deal out of human sexuality, and that she was surprised and "pleasantly amused" when she learned of her indirect contribution to image processing technology.

I'm just pointing out that the photo is more than the headshot. My culture doesn't make a big deal out of human sexuality either, and, as for the sexism issue, I'm not touching that with a ten foot pole on HN, because any opinion with the word "woman" in it somehow turns out to be sexist (probably including this one).

> I'm just pointing out that the photo is more than the headshot.

Not the photo that's actually used. Every photo is cropped, and a lot of great photos were cropped after they were taken. So it is with the photo of her used in the computer graphics world.

Just to make it clearer: There's a valid reason a photo of a human being showing some skin is used. Humans know, at a very deep level, what skin is supposed to look like. If the colors of skin are somewhat wrong, we'll know. We'll also know if the shape of the face is somewhat wrong, for example, which makes it easier to see other kinds of errors or infelicitous compression artifacts. All of this is further helped by using the same image in a wide variety of contexts over a long period of time.

> Not the photo that's actually used.

I've seen the full photo used, but that's probably because we don't really have a problem with nudes, and it's an iconic image.

I want computer vision's culture to not be a subset of cultures that are sexually open. I'm not personally offended, but I don't think we should risk turning away the contributions to computer vision from people who are slighted by the use of pornography in a scientific context.

She's beautiful. What's so wrong about appreciating someone's body? You don't see this kind of reaction to nude male models - that is where the real sexism is, IMO.

Imagery of nude male models (or sexualized masculinity in general) rarely features in material related to computer science or software development, so a lack of reaction in this context is to be expected.

Just as a sanity check: you are aware that graphics texts and papers don't use the entire Playboy photo, correct? I wonder if we're arguing about the same thing.

Point still stands that the material used in the software development field has no nudity in it what-so-ever. So it's a crop from a nude, big deal. Is it also wrong if a non-nude image of a porn actress is used? What about a non-nude of a regular woman? She's been naked at some point, after all.

Why is the fact that it's from a nude shot considered demeaning? Because it's from a porno magazine? So what? I could kind of understand if it was a segment of an image from a hardcore scene but the full photo is actually quite tasteful.

If I ever get the opportunity, I'll use a crop from a nude male photo to prove my point. I would bet a year's salary that there would be no backlash.

There are far bigger issues in the world to be indignant about than this. We have people dying of thirst and hunger when we technically have enough resources to feed every person on the planet yet there are individuals full of anger regarding trivialities like this where everything is consensual, nobody was hurt and nothing bad happened.

I believe that the correct term for this is "theatre of the absurd".

"There are far bigger issues in the world to be indignant about than this. We have people dying of thirst and hunger when we technically have enough resources to feed every person on the planet yet there are individuals full of anger regarding trivialities like this where everything is consensual, nobody was hurt and nothing bad happened."

By this logic, one could never discuss anything other than poverty, famine, and other terrible plights upon humanity.

Yet here we are, talking about such a variety of different topics.

I'm inclined to agree with you here. Unlike, for example, the timeless beauty of the Utah teapot, continuing to shoehorn Lena into digital image processing teaching and research does seem rather inappropriate in this day and age.

If you are willing to go a mile out of your way to look for things to be offended by, then, perhaps.

Not really, the issue is more about making computer science and related fields equally accessible and comfortable to all genders. One way to do this is to avoid unnecessarily biasing topics towards the sexual interests of heterosexual males, e.g. using such pornography in image processing materials.

I'm not personally offended by the image, but I don't think it's useful in the context just described.

First of all, most people in the field were unaware that the classical "Lenna" headshot came from Playboy, or from "pornography" in general. You may be overlooking the fact that some cultures don't consider her fully-nude photograph pornographic at all, while other cultures consider even the headshot pornographic. "Pornography" is not a very useful term of art... any art.

Second, women researchers are more than welcome to choose their own reference images.

Third, you aren't just condemning computer graphics researchers, but practitioners in fields from advertising to art to athletics to zymurgy. The female face (and form) has been celebrated for aesthetic and artistic purposes in every civilization this side of Afghanistan.

In short, I don't think I want to live in your world. I hope I continue to have a choice.

A world where female computer scientists and software developers feel as welcome as males sounds a lot better than the actual one plagued by rampant sexism and its corresponding denial.

For a few years I was righteously indignant about the gender issues involved but now it's just trite

Why should anyone's opinion matter except for Lenna's own? You are complaining about an uncoerced act.

For reference: http://www.cs.cmu.edu/~chuck/lennapg/lenna.shtml

Lena's opinion on the matter is not really the point - the criticism is along the same principle as exemplified here: http://martinfowler.com/bliki/SmutOnRails.html

It's at least as important as anyone else's.

I am pleased to hear that Lenna is personally OK with the situation, however I am more concerned with the thousands of young women attempting to enter the boys club of computer science. These women are being coerced to deal with pornography as a theme or abandon computer vision.

Reply on news.ycombinator.com