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

Drag Your GAN: Interactive Point-Based Manipulation of Images

vcai.mpi-inf.mpg.de

Post reply on HN
Print view View original

71–80 of 90 posts

Re: Drag Your GAN: Interactive Point-Based Manipulation of Images

#71

Post by vhcr »

nine_k wrote: post #56
vhcr wrote: post #46

Earlier quoted context omitted.

There's no way this would work, either the master encryption key would be leaked, or someone would reverse-engineer the chip. Also, what about someone putting a screen just in front of the sensor?

No need for that. A private key is generated on device and never leaves it. It sits inside a TPM or equivalent. The public key is pushed to a well-known site, visible to all. Every shot is signed by hashing the bits into a reasonably short string (say, using sha512) and then encrypted with the private key. Anyone can now decrypt the hash, and compare it with the hash they computed from the bits. The problem, of cours…

Why would you trust the key if it was generated on the device? Anyone could generate a key. You would either need a centralized server to sign that key, in which case you have to trust a centralized company / government agency.

Re: Drag Your GAN: Interactive Point-Based Manipulation of Images

#72

Post by blueyoda »

ftufek wrote: post #40
tikkun wrote: post #22

Has anyone built the "online photoshop that incorporates all of the latest AI image editing tools asap and sells access as a premium subscription with lots of GPU access for smooth editing" business yet? I'd be curious to know.

I think less effort has gone into image editing compared to image generation so far. That said, we're building some photo realistic image editing tools at https://www.faceshape.com , focused on face editing for now. Current models don't perform as well, but next generation currently under training will. I'm always curious to know what kind of AI image editing people are interested in, can you share what kind of edits…

Tried your app, I liked the product idea but think the execution could use much more work. Personally, I am obsessed with FaceApp’s filters. If you could make an app with even more interesting filters but with the same (or better) realistic quality, I’d definitely use it :)

Re: Drag Your GAN: Interactive Point-Based Manipulation of Images

#73

Post by qingcharles »

krunck wrote: post #41

I'm thinking that soon video and images are going to be just dead weight in journalism, adding nothing other than decoration.

Why just the video and images? Half of the stories on the Internet are created out of whole cloth by AI now, never mind the embedded media...

Re: Drag Your GAN: Interactive Point-Based Manipulation of Images

#74

Post by dragonwriter »

nine_k wrote: post #5

Hello, post-truth world! More seriously, I think that digital photos, and especially low-res surveillance camera coverage, will soon be inadmissible in any reasonable court, because tools like this would allow to forge such evidence in very natural-looking ways.

You have always been able to forge photographic evidence in natural looking ways (most easily by using completely real photography with misrepresented time or context).

And, of course, the easiest to falsify evidence (eyewitness testimony) is still admissible.

That's why you have to provide support for provenance, all of which is subject to examination and counterevidence.

Re: Drag Your GAN: Interactive Point-Based Manipulation of Images

#75

Post by nadermx »

vaidhy wrote: post #66
echelon wrote: post #35

Earlier quoted context omitted.

There are a million of these. It's a super crowded space. https://civitai.com/ https://lexica.art/ https://openart.ai/ (Many more)

They are all for image creation. I would love to have one to edit my photos.. not generate images from prompt.

https://ImageEditor.ai

Re: Drag Your GAN: Interactive Point-Based Manipulation of Images

#76

Post by flangola7 »

belter wrote: post #47
Giorgi wrote: post #36

Earlier quoted context omitted.

Adobe AI is a crap compared to Midjourney

Because one is on proper stock art and the other on anything without asking the creators for authorization? "AI art tools Stable Diffusion and Midjourney targeted with copyright lawsuit" - https://www.theverge.com/2023/1/16/23557098/generative-ai-ar...

Search on Twitter, the artists in the Adobe dataset are also livid (justifiably) because they didn't consent to their work being used for AI training. The Adobe license agreement is broad enough that Adobe is covered legally, but it isn't enthusiastic consent in any sense of the word. Many, many artists would never have submitted their work to Adobe if generative AI had been a known possibility, so using Adobe's product is really not any better at respecting creators' wishes.

Re: Drag Your GAN: Interactive Point-Based Manipulation of Images

#77

Post by tehwebguy »

grumbel wrote: post #59
nine_k wrote: post #5

Hello, post-truth world! More seriously, I think that digital photos, and especially low-res surveillance camera coverage, will soon be inadmissible in any reasonable court, because tools like this would allow to forge such evidence in very natural-looking ways.

Being able to fake something really doesn't matter all that much, as you'd still need to get that fake video into the surveillance camera system and you need to do so in the time between committing the crime and the police arriving, and without leaving a trace and hoping that whoever you try to incriminate doesn't have an alibi. Fakes will be relevant for Twitter, TikTok and Co., where random videos are posted and di…

Without knowing what the current ones are, I'd guess we need improved chain-of-custody laws for evidence immediately!

Re: Drag Your GAN: Interactive Point-Based Manipulation of Images

#78

Post by xrd »

tikkun wrote: post #22

Has anyone built the "online photoshop that incorporates all of the latest AI image editing tools asap and sells access as a premium subscription with lots of GPU access for smooth editing" business yet? I'd be curious to know.

I like the direction of github.com/invoke-ai/. It isn't a business but an open source project.

Re: Drag Your GAN: Interactive Point-Based Manipulation of Images

#79

Post by Der_Einzige »

Giorgi wrote: post #36
jahewson wrote: post #31

Earlier quoted context omitted.

Adobe Firefly already did it https://www.adobe.com/sensei/generative-ai/firefly.html

Adobe AI is a crap compared to Midjourney

Midjourney is crap compared to stable diffusion with all the features in automatic1111

Re: Drag Your GAN: Interactive Point-Based Manipulation of Images

#80

Post by u385639 »

famouswaffles wrote: post #70
u385639 wrote: post #68

Earlier quoted context omitted.

I'm not a critic of AI or moving any goal posts. I'm not lobbing comments in a vacuum. I was responding directly to the comical proposal that we don't need actors anymore, to which my Godfather comment has every relevance. Thanks anyway!

Guess i just don't think your comment has as much relevance as you think it does. Remove the "with ai" and nothing actually changes. "Please try to make an original movie that meets the standard of The Godfather, without AI" and lets see how well that goes. Is the human that fails this task also missing the "human element" ?

https://en.wikipedia.org/wiki/Rule_of_inference
Post reply on HN