Springer Nature retracts nearly 40 publications that trained neural networks on dataset containing images of children’s faces downloaded from websites about autism (www.thetransmitter.org)

submitted 1 day ago by spit_evil_olive_tips@beehaw.org to c/science@beehaw.org

11 comments fedilink hide all child comments

top 11 comments

sorted by: hot top controversial new old

[-] spit_evil_olive_tips@beehaw.org 23 points 1 day ago

I'm sort of speechless at how mind-bogglingly stupid every step of this process has been:

The papers attempted to train neural networks to distinguish between autistic and non-autistic children in a dataset containing photos of children’s faces. Retired engineer Gerald Piosenka created the dataset in 2019 by downloading photos of children from “websites devoted to the subject of autism,” according to a description of the dataset’s methods, and uploaded it to Kaggle, a site owned by Google that hosts public datasets for machine-learning practitioners.

The dataset contains more than 2,900 photos of children’s faces, half of which are labeled as autistic and the other half as not autistic.

After learning about a paper that cites the dataset, “I went and downloaded the dataset, and I was completely horrified,” says Dorothy Bishop, emeritus professor of developmental neuropsychology at the University of Oxford. “When I saw how it was created, I just thought, ‘This is absolute bonkers.’”

Without identifying each child in the dataset, there is no way to confirm that any of them do or do not have autism, Bishop says.

[-] supersquirrel@sopuli.xyz 18 points 1 day ago

This is disgusting, especially in this moment I am honestly repulsed someone would look for a way to visually target and identify autistic people, there is no kind reason to seek this capability when so many people desire to hurt autistic people.

[-] bmaxv@noc.social 10 points 1 day ago

@supersquirrel @spit_evil_olive_tips

Idk about stupid.

It's the direct consequence of having and collecting a bunch of data. Being able to come up with any idea, throw it against the wall and see what sticks was the entire point in the first place. Having almost all of them fail to find the ones that don't was the point of the setup.

This kind of use is also the reason why anyone who ever warned about data collection warned about it.

[-] supersquirrel@sopuli.xyz 8 points 1 day ago

My point is if someone handed me a tool that could decisively identify autistic kids from photographs I would destroy it if possible and then and advocate for parents never uploading photos of their autistic kids to the internet.

There are people in power who would immediately use this tool to do evil things, to seek to create it makes it clear the creators clearly don't give a shit about any other goal than "I figured something out!" in their research.

[-] bmaxv@noc.social 4 points 1 day ago

@supersquirrel I agree completely. The intent, setup, participation (including springer nature and everyone who publishes there), isn't "dumb" though.

It's malicious, evil, negligent or whatever you want to call it.

I dislike "stupid" because it leaves that room for innocent mistakes and unintentional behavior.

People didn't care for ethical standards and this is the outcome.

[-] FaceDeer@fedia.io 3 points 1 day ago

The best current treatments for autism depend on knowing as early as possible that a child needs specialized help to have the best outcomes. If there was a tool that decisively identified autistic kids and you destroyed it I would rank you as a particularly heinous monster.

[-] Bougie_Birdie@piefed.blahaj.zone 2 points 1 day ago

This tool wouldn't work very well either. Ignoring the problems of collecting and verifying the validity of the data (and those are pretty big, serious problems), 2,900 images isn't really enough to train an accurate image classifier. Especially not one that I would be comfortable using in a medical context.

Using the tool would probably result in a ton of false positives, and I'd bet the model would be overfitted to its data. Of course, I don't think that would prevent people from using something like that nefariously.

[-] MountingSuspicion@reddthat.com 1 points 1 day ago

While I agree that especially in this moment this is a terrible idea, and the collection process is idiotic and reprehensible, in a better society it might be helpful if it were actually a successful tool. Autism is a spectrum and a somewhat vague collection of behavioral traits and plenty of children grow up without a diagnosis causing lots of undue pain and stress. Plenty of them may learn to mask on their own, but imho it would be great to be able to catch it early and give them tools that might help them succeed and give their family more understanding of them as individuals. Again, I don't think this is a good idea, but I can see how a well meaning person might have come up with this.

[-] calliope@retrolemmy.com 9 points 1 day ago* (last edited 1 day ago)

I like this part. Someone brought it up to this guy

A Kaggle user broached these same concerns in a comment, and Piosenka responded that he had not violated privacy restrictions because all images were publicly available, adding “how can one be more ethical than to try to foster early detection and treatment of Autism in children. You sir are way off base.”

Yeesh. “How can anyone be more ethical than what I’m doing?? Not possible.”

That’s embarrassing!

[-] Paradachshund@lemmy.today 6 points 1 day ago

This is never how copyright has worked, and that's not even talking about getting permission from the person the photo is of. 😑

Why do so many AI people think that just because you found a way to scrape something, it's OK to use however you want?

[-] jherazob@beehaw.org 3 points 13 hours ago

Because one hallmark of the techbro mentality is not understanding consent

this post was submitted on 09 Dec 2025

50 points (98.1% liked)

Science

14790 readers

19 users here now

Studies, research findings, and interesting tidbits from the ever-expanding scientific world.

Subcommunities on Beehaw:

Space

Be sure to also check out these other Fediverse science communities:

This community's icon was made by Aaron Schneider, under the CC-BY-NC-SA 4.0 license.

founded 3 years ago

MODERATORS

alyaza@beehaw.org

Gaywallet@beehaw.org

remington@beehaw.org