I'm thinking that soon video and images are going to be just dead weight in journalism, adding nothing other than decoration.
Drag Your GAN: Interactive Point-Based Manipulation of Images
41–50 of 90 posts
Re: Drag Your GAN: Interactive Point-Based Manipulation of Images
#42I'm thinking that soon video and images are going to be just dead weight in journalism, adding nothing other than decoration.
It seems to be a pretty common thing now on news network websites, that at the top of the story, or perhaps somewhere in the middle, there is a video. But, you click on the video, and it's something entirely unrelated to the article. I feel like this as been going on for quite a while, nothing to do with synthetic media, but just an observation of an annoying pattern I've noticed.
Re: Drag Your GAN: Interactive Point-Based Manipulation of Images
#43I'm thinking that soon video and images are going to be just dead weight in journalism, adding nothing other than decoration.
I think there has to be some kind of cryptographic signature solution, like "The BBC verifies that this image is authentic".
Re: Drag Your GAN: Interactive Point-Based Manipulation of Images
#44The only thing that's new is the interactive interface the rest of it is old tech... You can use it on art breeder.com. Photoshop even has it in their face neural filter. GANs are not feasible for a variety of reasons you need to have models specific to your subject ie. why they change to a elephant model to manipulate the elephant.they are also not style agnostic but it's a great demo and the right time to release it. Just before the summit of the hype curve I bet one VC is dumb enough to throw millions at them.
Re: Drag Your GAN: Interactive Point-Based Manipulation of Images
#45The only thing that's new is the interactive interface the rest of it is old tech... You can use it on art breeder.com. Photoshop even has it in their face neural filter. GANs are not feasible for a variety of reasons you need to have models specific to your subject ie. why they change to a elephant model to manipulate the elephant.they are also not style agnostic but it's a great demo and the right time to release i…
With ChatGPT, the only thing new was the chat interface. In fact even Sam Altman mentioned this on Lex Fridman's podcast, IIRC - he said what he was most surprised about was the outsized effect the interface had on bringing LLM to the forefront of public consciousness, despite the existing maturity of the underlying GPT models. At least in that case it was OpenAI adding interactivity to its own existing models. But similarly, from a more holistic viewpoint, OpenAI productized existing research from Google. Transformer models were "old tech" since Google published "Attention is all you need" in 2017... and yet, when OpenAI managed to turn it into a usable product, suddenly they became the first movers and the company to beat. So I'm not convinced that only a "dumb" investor would fund an effort with a proven ability to productize "old tech."
Re: Drag Your GAN: Interactive Point-Based Manipulation of Images
#46Hello, 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.
There are probably ways to embed cryptographic hashes within images. Any device that creates images from the real world could have secret keys that can be used to validate any image created by said device. We will still need a centralized party that holds the secret keys for validation through
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?
Re: Drag Your GAN: Interactive Point-Based Manipulation of Images
#47Earlier 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
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...
Re: Drag Your GAN: Interactive Point-Based Manipulation of Images
#48The only thing that's new is the interactive interface the rest of it is old tech... You can use it on art breeder.com. Photoshop even has it in their face neural filter. GANs are not feasible for a variety of reasons you need to have models specific to your subject ie. why they change to a elephant model to manipulate the elephant.they are also not style agnostic but it's a great demo and the right time to release i…
This reads a lot like 'dropbox is trivial, rsync already exists'
Re: Drag Your GAN: Interactive Point-Based Manipulation of Images
#49What do we need human actors for at this point? Everything can be generated by AI now.
> What do we need human actors for at this point? Everything can be generated by AI now.
For blockbusters? If the tech still isn't there, it may be soon, and we won't need actors.
For cinema where we care it was made by humans, for humans? Actors will always be needed. Also, theater still exists and people enjoy it.
Re: Drag Your GAN: Interactive Point-Based Manipulation of Images
#50The only thing that's new is the interactive interface the rest of it is old tech... You can use it on art breeder.com. Photoshop even has it in their face neural filter. GANs are not feasible for a variety of reasons you need to have models specific to your subject ie. why they change to a elephant model to manipulate the elephant.they are also not style agnostic but it's a great demo and the right time to release i…
This reads a lot like 'dropbox is trivial, rsync already exists'
[flagged]