3 ms·

Why do you believe that?

melissalobos 4y ago
It was mentioned in the DALL-E github repo.
egeozcan 4y ago
Coming up with ok-looking generated faces is not ground-breaking in the ML world and when the state-of-the-art model botches it completely, you start looking for the reason.
- lupire 4y ago
  And coming up with very good looking modifications/poses of known actual faces is very well achieved.
- shadowgovt 4y ago
  How compose-able are ML solutions? I'm wondering if coming up with ok-looking generated headshots (face at a specific, known, predetermined angle, in a neutral and standardized context, independent of background features) is not ground-breaking, but embedding that capability in a dynamic image generator with myriad other objects, headgear, lights and shadows, etc. may be.
  - gwern 4y ago
    Facebook's Make-A-Scene, which looks almost as good as DALL-E 2, does something very much like that, using face-specific targeting to boost the quality of that part of images: https://arxiv.org/abs/2203.13131#facebook https://arxiv.org/abs/2203.13131#facebook
nullc 4y ago
Because they said so: Preventing Harmful Generations We’ve limited the ability for DALL·E 2 to generate violent, hate, or adult images. By removing the most explicit content from the training data, we minimized DALL·E 2’s exposure to these concepts. We also used advanced techniques to prevent photorealistic generations of real individuals’ faces, including those of public figures.
- vmception 4y ago
  Is there a fork that unhandicaps it?
  - not2b 4y ago
    To do such a fork, wouldn't you have to build a completely new model with the same training data and everything else the same except leaving out the restrictions?
    - somebodythere 4y ago
      You should be able to finetune on a dataset that has captions of human faces.
    - heavyset_go 4y ago
      No, you can use transfer learning.
  - nullc 4y ago
    Like a lot of other Open AI work, DALL-E 2 isn't open. There are people working at re-implementing it, but training has considerable computational costs.

It was mentioned in the DALL-E github repo.

Coming up with ok-looking generated faces is not ground-breaking in the ML world and when the state-of-the-art model botches it completely, you start looking for the reason.

lupire 4y ago
And coming up with very good looking modifications/poses of known actual faces is very well achieved.
shadowgovt 4y ago
How compose-able are ML solutions? I'm wondering if coming up with ok-looking generated headshots (face at a specific, known, predetermined angle, in a neutral and standardized context, independent of background features) is not ground-breaking, but embedding that capability in a dynamic image generator with myriad other objects, headgear, lights and shadows, etc. may be.
- gwern 4y ago
  Facebook's Make-A-Scene, which looks almost as good as DALL-E 2, does something very much like that, using face-specific targeting to boost the quality of that part of images: https://arxiv.org/abs/2203.13131#facebook https://arxiv.org/abs/2203.13131#facebook

nullc 4y ago

Because they said so: Preventing Harmful Generations We’ve limited the ability for DALL·E 2 to generate violent, hate, or adult images. By removing the most explicit content from the training data, we minimized DALL·E 2’s exposure to these concepts. We also used advanced techniques to prevent photorealistic generations of real individuals’ faces, including those of public figures.

vmception 4y ago
Is there a fork that unhandicaps it?
- not2b 4y ago
  To do such a fork, wouldn't you have to build a completely new model with the same training data and everything else the same except leaving out the restrictions?
  - somebodythere 4y ago
    You should be able to finetune on a dataset that has captions of human faces.
  - heavyset_go 4y ago
    No, you can use transfer learning.
- nullc 4y ago
  Like a lot of other Open AI work, DALL-E 2 isn't open. There are people working at re-implementing it, but training has considerable computational costs.