2 ms·

I think what it comes down to is that most of the other ways to automate ID creation at scale were probably really easy to detect using ... well, probably neura

I think what it comes down to is that most of the other ways to automate ID creation at scale were probably really easy to detect using ... well, probably neural networks. Because ID verification services had access to millions of "real" ID photos, I'd think a lot of fake images created using traditional automation would be pretty easy to detect (For example, having the same backgrounds/wood grain pattern reused for multiple images). So what neural network generation is adding here is basically being able to add much much more "plausible randomness" to the image (lens artifacts, dust motes, wood grain, plastic reflection, etc) in a way that makes distinguishing generated from non-generated images statistically very difficult.

jszymborski 3y ago
I'm not going to give these folks too many details, but there have long been ways to create convincing synthetic organic textures that predate NNs. Although I think you're right, the market tells us that this is likely either too high-skill and/or high-labour and/or easily detected.
Muromec 3y ago
> Because ID verification services had access to millions of "real" ID photos It's all snake oil devised to shift the risk from a party who has a mandate to perform KYC to some broken by design AI woodoo, which will generate free money until it doesn't once the landscape or the legislator's will mood change. They bank will then perform the actual KYC and will start asking questions once it sees something funny and will eat the cost of fraud at the end of the day.