this post was submitted on 22 Jan 2024
142 points (98.6% liked)
Technology
58137 readers
4359 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related content.
- Be excellent to each another!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, to ask if your bot can be added please contact us.
- Check for duplicates before posting, duplicates may be removed
Approved Bots
founded 1 year ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
IMO, the only valid use of DNA-based face generation would be to rule out existing suspects, not to label random people as suspects based on their faces alone.
I hate to say it, but there’s a better way to eliminate suspects based on our current DNA technology.
If all your suspects are black, and the dna is from a someone with Irish heritage, it’s probably not any of the black people.
Trying to reconstruct someone’s face seems really inaccurate, considering I have the same DNA as I did 10 years ago, but I’ve had high school friends who have walked passed me without recognizing me because I lost a lot of weight and grew a beard since they last saw me.
As much as racial profiling is shitty, it’s way easier to tell someone’s ethnicity from dna than it is to reconstruct their whole face. You can then use that to narrow down a list of suspects, similar to how we used to use blood type analysis before dna was a thing.
While I'm as skeptical as you are, I don't think people recognizing you is a good metric.
A better test would be if an AI trained on your younger face could accurately and reliably identify you with your adult face.
The way AI and human face recognition work are different from each other. An AI may be able to identify you based on markers that human recognition doesn't account for
AI does a scary good job of recognizing age different photos sometimes. I've set up a couple self hosted photo management apps that contain such functions (photo prism and immich) and had surprising results. After feeding it a number of recent digital pictures I went on to put on n a bunch of old scanned photos from 10+ years earlier, and with a reasonably good accuracy it was able to dicern the difference in baby pictures between two kids to match with the older pictures.