Posts by BlackPhi@geekdom.social
(DIR) Post #Ac77GnpnbeakH4hCQy by BlackPhi@geekdom.social
2023-11-23T23:26:26Z
0 likes, 0 repeats
@john The thing about Stable Diffusion is that it is not a tool for following a logical set of instructions. It is about associations and links into features of existing images on the internet. In the case of your prompt, the associations attached to 'autumn' and 'leaves' override the associations of 'terraced houses' and 'residential street'. Also, pictures on the internet of foxes with human faces are, as you say above, not common, so SD doesn't have a lot to work from.
(DIR) Post #Ac77GovrWb2HgAXYWW by BlackPhi@geekdom.social
2023-11-23T23:37:37Z
0 likes, 0 repeats
@john Another thing about SD is that it can be sensitive to your prompt details. There may be few foxes with human faces on the internet, but there are plenty of foxes so it should be able to do a lot better than your outputs. The challenge of using SD as a tool can often be in getting the feel for how it likes to work and going with it - more like sailing than motoring. Then if you get a picture you like with a fox, SD lets you experiment with infilling the face area with a new prompt.
(DIR) Post #Ac78cVdA4II28ohOHw by BlackPhi@geekdom.social
2023-11-24T00:01:59Z
0 likes, 0 repeats
@john Basically SD 'knows' about faces because it has 'seen' millions of images tagged as being faces- almost all of them human. It probably has very little reference to fox faces, other than in children's drawings, so it pretty much ignores that part of your prompt.Dall-E is in many ways a rather different beast to SD: easier to control perhaps, but less inclined to come up with interesting and original 'ideas', I think. It's a bit like watercolour vs oil paint, perhaps.
(DIR) Post #Ac85XZdfdJ5kQ6qsm8 by BlackPhi@geekdom.social
2023-11-24T11:02:06Z
0 likes, 0 repeats
@john Something like this? The trickiest bit was getting it so it doesn't look photoshopped. The inpainting was denoised at 0.9. I don't know if the Americanisations made any difference, I suspect not. There is cleaning up to do around the tail and legs, but that is to be expected with #stablediffusion. The initial model was SD's 768-v-ema and the inpainting used the Realistic Vision 5.1 inpainting model, masking the eyes and muzzle.
(DIR) Post #AjFYwxbE6909JEqQMa by BlackPhi@geekdom.social
2024-06-24T10:07:31Z
0 likes, 1 repeats
@ned Except that creating an image with Stable Diffusion takes around 4kJ on my PC whereas air conditioning a smallish flat for an hour takes nearer 4,000kJ (obviously rough approximations, but the right order of magnitude). See https://learnmetrics.com/how-many-kw-air-conditioner-do-i-need-ac-kw-calculator-chart/ for some approximations and note that 1kW*h is around 3,600kJ.There is a certain amount of misinformation and confusion around AI energy usage, so its always worth doing a rough sense check on such comparisons. #AI #AIArt #StableDiffusion
(DIR) Post #At2xR1mKpy4dvBnWiG by BlackPhi@geekdom.social
2025-04-13T11:23:39Z
0 likes, 0 repeats
@johnmacintosh Wow, that's a remarkable sequence of pictures!