Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Katie Bouman, the computer scientist behind the first black hole image

bbc.com

Post reply on HN
Print view View original

271–280 of 547 posts

Re: Katie Bouman, the computer scientist behind the first black hole image

#271

Post by jdmichal »

jamestenglish wrote: post #263
QuotedForTruth wrote: post #254

Earlier quoted context omitted.

Disclaimer: this is based on watching the talk and some basic machine learning knowledge. Im no expert. They have a sparse set of data that is part of an image. They have trained a model to look at the sparse set and make an educated guess about what the full image looks like. They do this by feeding it full images. The full images you feed into the model thus have an effect on the final image generated. In order to…

Not OP, but I too am confused. I understood the sketch artist analogy but that didn't seem related to this point: >They also trained the model with non-blackhole images. Since the output of the model was approximately the same, this indicates that the resulting output picture doesnt look like what we think a black hole looks like just because it was trained with black hole images. It likely really looks like that. If…

They're not just training the model to make pictures from nothing. They're training the model to make pictures from an input.

So I assume they're simulating what an input would look like of, say, a planet or astroid or elephant or whatever, given that it was viewed through the relevant type of sensor system. Then when they feed in the black hole sensor data, they get pictures that look like the black holes we imagined. Even if we never told the model what a black hole looks like.

Re: Katie Bouman, the computer scientist behind the first black hole image

#272

Post by dvtrn »

princeofwands wrote: post #269
dvtrn wrote: post #251

Earlier quoted context omitted.

Can I ask how you know what I would or wouldn’t do and what my sensibilities are on this topic?

A most favorable reading of your post has you: - dragging in gender by the hairs (as if it is relevant) - trying out the trope that people are denying her title, because she is female. If anything, I see you contributing to sexism, even though I believe your intentions to white knight are good (though likely biologically inspired to enhance your chance of mating).

I’m not white knighting anything. I am respecting her academic achievements as someone raised by someone who holds a doctorate to respect and address people by the titles they’ve earned-where appropriate, and encouraging others to do the same. I've literally said nothing about what bearing her gender has on my opinion--I only offered the opinion that the article's headline should acknowledge the title this individual, one who happens to be a woman, holds in light of their remarkable scientific achievement.

Why is this a problem and why do you characterize respecting achievements as “white knighting”?

Are you here to tell me a man can’t respect a woman’s title without having an ulterior motive? Why do you believe this to be the case?

Re: Katie Bouman, the computer scientist behind the first black hole image

#273

Post by QuotedForTruth »

jamestenglish wrote: post #263
QuotedForTruth wrote: post #254

Earlier quoted context omitted.

Disclaimer: this is based on watching the talk and some basic machine learning knowledge. Im no expert. They have a sparse set of data that is part of an image. They have trained a model to look at the sparse set and make an educated guess about what the full image looks like. They do this by feeding it full images. The full images you feed into the model thus have an effect on the final image generated. In order to…

Not OP, but I too am confused. I understood the sketch artist analogy but that didn't seem related to this point: >They also trained the model with non-blackhole images. Since the output of the model was approximately the same, this indicates that the resulting output picture doesnt look like what we think a black hole looks like just because it was trained with black hole images. It likely really looks like that. If…

This is ensuring that the model is not over trained.

They also showed that when they fed in simulated sparse measurements based on real full images of generic things, they got back fuzzy versions of the real image. [1] So if you put in a sparsely captured elephant (if for instance there was one at the center of the galaxy) you'd get an image of the elephant out, not this black hole.

To complete the artist analogy, imagine that the suspect that is being drawn by each artist is some stereotypical American. The description given to the artists doesnt say that, it just describes how the person looks. One of the three sketch artists is American and the others are Chinese and Ethiopian.

If the American draws a stereotypical American, how can you be sure that the drawing is accurate and thats not just what he assumed the person would look like because everyone he has ever seen looks like that?

You look at what the other two draw. If they both draw the same stereotypical American, even though they have no knowledge of what a stereotypical American looks like, you can be pretty sure that they determined that based on the description provided to them. The actual data.

They did still likely utilize some of their knowledge about what humans in general look like though. This is analogous to how the model uses its training on what a generic image looks like. For instance, maybe several sparse pixels of the same value are likely to have pixels of that same value between them. The model puts things like this together and spits out a picture of what we think a black hole looks like even though its never seen a black hole before.

[1] https://youtu.be/BIvezCVcsYs?t=685

Re: Katie Bouman, the computer scientist behind the first black hole image

#274

Post by govg »

gonational wrote: post #262

Is this the correct repo? https://github.com/achael/eht-imaging/graphs/contributors I didn't realize this was public code. It looks like one "achael" is the author of this, though.

I don't think she was a large part of the implementation effort, especially with her background. Andrew Chael seems to have been the programmer on this project, or at least handled a lot of it. (Note a significant fraction of his code commits content are models)

Re: Katie Bouman, the computer scientist behind the first black hole image

#275

Post by nardi »

Gpetrium wrote: post #266

Reading some of the comments in here and in previous texts, I think everyone should try to heed to the following guidelines: - When an individual/team's work is emphasized by their biological characteristics, it is often meant for clicks or to drive emotions (positive & negative) - When that happens, ask yourself whether the author of the paper did it for nefarious reasons or not. - If the cause doesn't seem nefariou…

Seems to me that you are the one emphasizing “biological characteristics”. The article does not. Are you projecting, maybe?

Re: Katie Bouman, the computer scientist behind the first black hole image

#276

Post by stilley2 »

QuotedForTruth wrote: post #254
dennisgorelik wrote: post #242

Earlier quoted context omitted.

It is an interesting presentation, but I do NOT understand Katie's explanation about how they were going to minimize the bias [to "see" already predicted black hole visualization] while creatively interpreting inputs from sparsely placed telescopes around the earth. Do you understand Katie's explanation?

Disclaimer: this is based on watching the talk and some basic machine learning knowledge. Im no expert. They have a sparse set of data that is part of an image. They have trained a model to look at the sparse set and make an educated guess about what the full image looks like. They do this by feeding it full images. The full images you feed into the model thus have an effect on the final image generated. In order to…

Hmm I skimmed the paper on the algorithm this morning and didn't get the impression they trained the model on other images. I thought they jointly estimated patches that make up the image and penalized deviations from these patches (i.e. estimated a sparse basis). I haven't watched the Ted talk yet though

Re: Katie Bouman, the computer scientist behind the first black hole image

#277

Post by yeahitslikethat »

realradicalwash wrote: post #265

Her story is trully inspiring! She seems like a really likable person, has been hard working, with great results, making a major contribution. The photo with her and the hard drives is amazing and I am sure she will inspire many to enter science. However, I think to call her "the woman behind the first black hole image" is a hyperbole. It makes it sound as if she was _the one person_ responsible that all this came ab…

No one would read that story. No one knows Katie Bouman. Everyone knows the blackhole.

You're thinking like a programmer. Think not like a programmer and then explain why the original headline is better.

Why was the article written?

Re: Katie Bouman, the computer scientist behind the first black hole image

#278

Post by Gpetrium »

nardi wrote: post #275
Gpetrium wrote: post #266

Reading some of the comments in here and in previous texts, I think everyone should try to heed to the following guidelines: - When an individual/team's work is emphasized by their biological characteristics, it is often meant for clicks or to drive emotions (positive & negative) - When that happens, ask yourself whether the author of the paper did it for nefarious reasons or not. - If the cause doesn't seem nefariou…

Seems to me that you are the one emphasizing “biological characteristics”. The article does not. Are you projecting, maybe?

The Ycombinator text doesn't say it, but the link does. I think a lot of the negative reaction in here comes from the way people have started to perceive the narrative created by the media and personal feelings of threat (which is often unfounded).

I am all in for getting more people from all backgrounds into Computer & Sciences. I also agree that sometimes it is beneficial to have 'biological characteristics' added to articles to get certain groups to find someone to look up to. Humans are biologically set-up to do that, what the guideline implies is for everyone to do 'at least that' before creating biased comments.

Care to explain how the 'guidelines' comment is projecting?

Re: Katie Bouman, the computer scientist behind the first black hole image

#280

Post by dunslandsboo »

juliendorra wrote: post #103
dunslandsboo wrote: post #80

Earlier quoted context omitted.

Why do you want to change the behaviour of women regarding their choice of study field?

You misunderstood. We want any person, women and young girls included, to be able to pursue a career path, if they have even the faintest desire of it, without self-censorship, negative remarks, feeling out of place, their vocation and/or skills being continuously challenged randomly, or having to cope with various forms of harassment. If you build an environment that allow that, women presence in the field surge. An…

Maybe you can help me understand. So how do you explain why there are fewer women in STEM fields in Scandinavia and more in Turkey, Tunesia and United Arab Emirates?

Well at least according to the paper "The Gender-Equality Paradox in Science, Technology, Engineering, and Mathematics Education Gijsbert Stoet, David C. Geary"

https://journals.sagepub.com/doi/abs/10.1177/095679761774171...

Post reply on HN