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

Launch HN: Reality Defender (YC W22) – Deepfake Detection Platform

news.ycombinator.com

Post reply on HN
Print view View original

41–50 of 93 posts

Re: Launch HN: Reality Defender (YC W22) – Deepfake Detection Platform

#41

Post by prometheus76 »

smt88 wrote: post #27
cphoover wrote: post #5

I believe deep-fakes to be a serious threat to functioning democracies around the world. I would love to be on the front-lines fighting against this threat. I have submitted my resume: https://cphoover.github.io/ Have you also considered either offering a browser plugin to display contextual warnings attached to video elements? Or thought of working with browser makers? The web/social media is where a ton of fake med…

I worry about the same thing. I think "blockchain" (or even just a signed, immutable, public database) is mostly a solution in search of a problem, but I do think it may have an application here. If you can hash a video when it's recorded and publish the hash with a timestamp that can't be forged, you can at least prove that this video existed at least as long ago as that stamp. That allows you to invalidate any deep…

In that scenario, where is the line between a deepfake and satire?

Re: Launch HN: Reality Defender (YC W22) – Deepfake Detection Platform

#42

Post by calvinmorrison »

candiddevmike wrote: post #15
bpcrd wrote: post #11

Earlier quoted context omitted.

In 2017 deepfakes were pretty crude, today the avg person can’t tell a real face from a deepfake generated on a 5 year old iPhone. We expect the tech to continue moving in this direction. So, similar to anti-virus, we're approaching this problem with an iterative, multi-model solution that can evolve with the threat.

So the folks at the forefront of deep fake technology (i.e. the attackers you're targeting) will slip through your product because it lags behind the state of the art (like AV, which you said is the approach you're following), while innocent folks will be caught by it due to a new kafkaesque version of "prove you're not a bot" since you focus on reducing false negatives. Hopefully I can avoid companies using your pro…

Of course, because these companies are probably owned by the same people in the end that develop the DeepFake datasets, generating endless income from both sides.

It's like ADA Compliance lawsuits. I can't prove the AccessaBe or other "ADA Compliance" web tooling are generating these lawsuits, but their company would not exist without them. Why wouldn't they want more lawsuits?

Re: Launch HN: Reality Defender (YC W22) – Deepfake Detection Platform

#43

Post by ChefboyOG »

calvinmorrison wrote: post #42
candiddevmike wrote: post #15

Earlier quoted context omitted.

So the folks at the forefront of deep fake technology (i.e. the attackers you're targeting) will slip through your product because it lags behind the state of the art (like AV, which you said is the approach you're following), while innocent folks will be caught by it due to a new kafkaesque version of "prove you're not a bot" since you focus on reducing false negatives. Hopefully I can avoid companies using your pro…

Of course, because these companies are probably owned by the same people in the end that develop the DeepFake datasets, generating endless income from both sides. It's like ADA Compliance lawsuits. I can't prove the AccessaBe or other "ADA Compliance" web tooling are generating these lawsuits, but their company would not exist without them. Why wouldn't they want more lawsuits?

The majority of large, popular datasets in deep learning are curated and hosted by academics:

https://paperswithcode.com/task/deepfake-detection#datasets

Re: Launch HN: Reality Defender (YC W22) – Deepfake Detection Platform

#44

Post by ChefboyOG »

sigil wrote: post #39
davidweatherall wrote: post #38

Earlier quoted context omitted.

(Not OP) This would work for videos that people want to acknowledge as their own, but they could just tweet out they own it for the same affect. The issue this tackles is videos that people don't want to claim ownership of, e.g. if a video emerged of kicking a child, the politician can't say "I haven't signed it therefore it's not mine", instead we need tools like the above to be able to say, "this is faked, do not t…

Maybe in the interim we need deepfake detection tools, but I'm asking about the long run. Suppose signed AV takes off as described above. The public has been trained: "If the video doesn't have The King's Seal, it's not from The King."

I think the point is that yes, that seems like a good solution for verifying content that purports to be released by a certain creator, but it doesn't solve the problem of deep fakes for captured footage i.e. you can prove it isn't a video that you created, but you can't prove it isn't a video someone else took of you.

Re: Launch HN: Reality Defender (YC W22) – Deepfake Detection Platform

#45

Post by wanderer_ »

bpcrd wrote: post #14
candiddevmike wrote: post #9

Earlier quoted context omitted.

Why would you be more concerned about false negatives? Wouldn't false positives erode trust and value in your product, and considering the applications you're targeting, possibly open you up to lawsuits if you start accusing innocent people of being deepfakes (which, IMO, currently seems unlikely)?

We provide a probabilistic percentage result that is used by a trust and safety team to set limits (ie. flag or block content) so it is not a binary yes/no. We search for specific deepfake signatures and we explain what our results are identifying.

So... that percentage sort of 'return type' allows the people using your service to decide how aggressive they want to be? Smart. It also could possibly turn your service into more of a tool and less of something that someone could blame incorrect results on.

Re: Launch HN: Reality Defender (YC W22) – Deepfake Detection Platform

#46

Post by sigil »

ChefboyOG wrote: post #44
sigil wrote: post #39

Earlier quoted context omitted.

Maybe in the interim we need deepfake detection tools, but I'm asking about the long run. Suppose signed AV takes off as described above. The public has been trained: "If the video doesn't have The King's Seal, it's not from The King."

I think the point is that yes, that seems like a good solution for verifying content that purports to be released by a certain creator, but it doesn't solve the problem of deep fakes for captured footage i.e. you can prove it isn't a video that you created, but you can't prove it isn't a video someone else took of you.

That makes sense. But if signed AV takes off, the video someone else took of you and shared likely bears their seal. And audiences decide how much they trust that source – just like they look for a CNN / BBC / etc logo in the corner currently.

Re: Launch HN: Reality Defender (YC W22) – Deepfake Detection Platform

#47

Post by version_five »

This is definitely an interesting problem, and I can see a place for it.

I want to comment (in ignorance because I don't know the techniques you are using) that there is more to detecting fakes or "misinformation" than just the digital attributes of the data. Specifically, confirmation from multiple sources, reputation of sources, and above all, consistency with a world model of how people behave. For example, if there's a video of Biden backing Putin, you could dismiss it as fake regardless of video attributes.

I think (and have been criticized for saying, why I don't understand), that education and emphasizing critical thinking are the biggest counters to fakes, not learning to spot them in feature space. I believe that whatever you make, sanity checks need to be a part of it, and not just blind flagging of true / false based on the digital attributes.

Thanks!

Re: Launch HN: Reality Defender (YC W22) – Deepfake Detection Platform

#49

Post by atlasunshrugged »

This is super cool, you should chat with the folks in Estonia who had worked on developing Sentinel (https://thesentinel.ai/) which had a similar premise but ended up pivoting. I advised them for a bit, happy to chat too (email in bio).

Edit to include link to their website

Post reply on HN