návod na zprovoznění ai artu, ale i jazykových modelů doma a technické materiály: https://bit.ly/af-2023-ai-art-doma
- má přednáška o AI na AnimeFestu 2019
- https://www.youtube.com/watch?v=WWrhXgaoSWM&ab_channel=AnimefestCZ
- prezentace: https://drive.google.com/file/d/12jW7YbojmWCK6XPFqHpFyANgM6pHlY0X/view
- zdroje: https://gist.github.com/racinmat/b01d8cdb4f515a136644542dac19a5ea
- aktualizovaná prezentace, 2022: https://docs.google.com/presentation/d/11J_cCNyo7Yds93XcwqmAVN5PEtRZ46Aj/edit?usp=sharing&ouid=102632222370728552977&rtpof=true&sd=true
- aktualizované zdroje, 2022: https://gist.github.com/racinmat/2f988eef70372ff90679ef88f450b8b8
- turing test
- počet vrstev deep NN
- víc než 3 http://wiki.pathmind.com/neural-network#:~:text=More%20than%20three%20layers%20(including,qualifies%20as%20%E2%80%9Cdeep%E2%80%9D%20learning.
- 3 a víc https://stats.stackexchange.com/questions/229619/minimum-number-of-layers-in-a-deep-neural-network ale včetně vstupní vrstvy je to taky víc než 3
- AI art
- https://en.wikipedia.org/wiki/Artificial_intelligence_art
- AARON - první průkopník
- stylegan anime https://gwern.net/face
- atomů ve vesmíru https://www.thoughtco.com/number-of-atoms-in-the-universe-603795#:~:text=The%20universe%20is%20vast.,random%2C%20made%2Dup%20number.
- výpočet možností dalle 1 dvae https://www.wolframalpha.com/input?i=%3D8192%5E1024
- https://www.pathofexile.com/forum/view-thread/2138904
- https://civitai.com/images/626680?modelVersionId=56762&prioritizedUserIds=989837&period=AllTime&sort=Most+Reactions&limit=20
- čínská služba: fotka->anime obrázek
- deep dream
- https://en.wikipedia.org/wiki/DeepDream
- https://www3.cs.stonybrook.edu/~cse352/T12talk.pdf
- původní článek https://ai.googleblog.com/2015/07/deepdream-code-example-for-visualizing.html
- inception paper https://arxiv.org/pdf/1409.4842.pdf
- 2015 článek https://www.americanscientist.org/article/computer-vision-and-computer-hallucinations
- online na google colabu https://colab.research.google.com/github/tensorflow/docs/blob/master/site/en/tutorials/generative/deepdream.ipynb#scrollTo=F4RBFfIWNbG0
- this XYZ does not exist
- midjourney
- dohady, co používá https://nerdschalk.com/does-midjourney-use-stable-diffusion/
- dalle, dalle 2
- blogpost dalle 2 https://openai.com/product/dall-e-2
- blogpost dalle 1 https://openai.com/research/dall-e
- online demo https://labs.openai.com/
- dalle paper https://arxiv.org/pdf/2102.12092.pdf
- dalle dvae zdroják https://github.com/openai/DALL-E
- dalle 2 paper https://arxiv.org/pdf/2204.06125.pdf
- dalle blogpost https://openai.com/research/dall-e
- vysvětlení fungování
- VQ-VAE a DVAE
- vysvětlení fungování https://www.youtube.com/watch?v=VZFVUrYcig0
- paper https://arxiv.org/pdf/1711.00937.pdf
- článek analyzující dvae vocabulary https://towardsdatascience.com/almost-any-image-is-only-8k-vectors-c68c1b1aa6d2
- midjourney
- stable diffusion
- github se zdrojáky univerzitní https://github.com/CompVis/stable-diffusion
- runway ml větev https://github.com/runwayml/stable-diffusion
- web paperu https://ommer-lab.com/research/latent-diffusion-models/
- navazující výzkum https://ommer-lab.com/research/latent-diffusion-models/further-research-based-on-stable-diffusion/
- paper https://arxiv.org/pdf/2112.10752.pdf
- autoři jsou jak na univerzitě, tak ve stability.ai
- vztah stability.ai ke stable diffusion https://www.technologyreview.com/2022/09/16/1059598/this-artist-is-dominating-ai-generated-art-and-hes-not-happy-about-it/
- vztah stability ai, runway ml a univerzity https://futurism.com/the-byte/stable-diffusion-stability-ai-risk-going-under
- public release https://stability.ai/blog/stable-diffusion-public-release
- vysvětlení od computerphile https://www.youtube.com/watch?v=1CIpzeNxIhU&t=2s&ab_channel=Computerphile
- blog s vysvětlením https://jalammar.github.io/illustrated-stable-diffusion/
- video z fast.ai popisující důkladně principy https://www.youtube.com/watch?v=_7rMfsA24Ls
- https://medium.com/@steinsfu/stable-diffusion-clearly-explained-ed008044e07e
- valikosti datasetů
- lsun https://arxiv.org/pdf/1506.03365.pdf
- LAION
- have i been trained
- https://www.theverge.com/2023/1/16/23557098/generative-ai-art-copyright-legal-lawsuit-stable-diffusion-midjourney-deviantart
- browser laion aestehetic https://laion-aesthetic.datasette.io/laion-aesthetic-6pls/images
- https://80.lv/articles/exploring-the-images-used-to-train-stable-diffusion-s-ai/
- https://www.theverge.com/2023/1/16/23557098/generative-ai-art-copyright-legal-lawsuit-stable-diffusion-midjourney-deviantart
- https://www.theverge.com/2022/11/8/23446821/microsoft-openai-github-copilot-class-action-lawsuit-ai-copyright-violation-training-data
- https://www.theverge.com/2023/1/17/23558516/ai-art-copyright-stable-diffusion-getty-images-lawsuit
- https://analyticsindiamag.com/stability-ai-midjourney-face-legal-woes-as-shutterstock-settles-with-artists/
web ui: https://github.com/AUTOMATIC1111/stable-diffusion-webui
- DCGAN https://arxiv.org/abs/1511.06434
- CLIP
- blogpost https://openai.com/research/clip
- zdroják https://github.com/openai/CLIP
- paper https://arxiv.org/pdf/2103.00020.pdf
- VIT
-
GPT https://en.wikipedia.org/wiki/Generative_pre-trained_transformer
-
chatgpt
-
pygmalion ai
-
velikosti modelů:
- https://www.vinayiyengar.com/2022/08/04/the-promise-and-perils-of-large-language-models/
- https://huggingface.co/blog/large-language-models
- https://datascience.stackexchange.com/questions/117444/what-size-language-model-can-you-train-on-a-gpu-with-x-gb-of-memory
- https://github.com/Xirider/finetune-gpt2xl
- https://www.techtimes.com/articles/266510/20211011/nvidia-microsoft-megatron-language-processor.htm
- https://bdtechtalks.com/2020/09/21/gpt-3-economy-business-model/