You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We present high quality image synthesis results using diffusion probabilistic models, a class of latent variable models inspired by considerations from nonequilibrium thermodynamics. Our best results are obtained by training on a weighted variational bound designed according to a novel connection between diffusion probabilistic models and denoising score matching with Langevin dynamics, and our models naturally admit a progressive lossy decompression scheme that can be interpreted as a generalization of autoregressive decoding. On the unconditional CIFAR10 dataset, we obtain an Inception score of 9.46 and a state-of-the-art FID score of 3.17. On 256x256 LSUN, we obtain sample quality similar to ProgressiveGAN.
Sohl-Dickstein, Jascha, Eric Weiss, Niru Maheswaranathan, and Surya Ganguli. 07--09 Jul 2015. “Deep Unsupervised Learning Using Nonequilibrium Thermodynamics.” In Proceedings of the 32nd International Conference on Machine Learning, edited by Francis Bach and David Blei, 37:2256–65. Proceedings of Machine Learning Research. Lille, France: PMLR.
拡散確率モデルを用いた画像生成を提唱した論文。
Song, Yang, and Stefano Ermon. 2019. “Generative Modeling by Estimating Gradients of the Data Distribution.” arXiv [cs.LG]. arXiv. http://arxiv.org/abs/1907.05600.
Dhariwal, Prafulla, and Alex Nichol. 2021. “Diffusion Models Beat GANs on Image Synthesis.” arXiv [cs.LG]. arXiv. http://arxiv.org/abs/2105.05233.
言語モデルに進出
Austin, Jacob, Daniel D. Johnson, Jonathan Ho, Daniel Tarlow, and Rianne van den Berg. 2021. “Structured Denoising Diffusion Models in Discrete State-Spaces.” arXiv [cs.LG]. arXiv. http://arxiv.org/abs/2107.03006.
Ho, Jonathan, Ajay Jain, and Pieter Abbeel. 2020. “Denoising Diffusion Probabilistic Models.” arXiv [cs.LG]. arXiv. http://arxiv.org/abs/2006.11239.
Revised 2020
Abstract
(DeepL翻訳)
我々は、非平衡熱力学の考察に触発された潜在変数モデルのクラスである拡散確率モデルを用いた高品質な画像合成の結果を発表する。我々の最良の結果は、拡散確率モデルとランジュバン動力学によるノイズ除去のスコアマッチングとの間の新しい接続に従って設計された重み付き変分境界で学習することによって得られ、我々のモデルは自然に自己回帰復号の一般化として解釈できる漸進的損失伸長方式を認める。無条件CIFAR10データセットにおいて、Inceptionスコア9.46、FIDスコア3.17を得ることができました。256x256 LSUNでは、ProgressiveGANと同程度のサンプル品質が得られている。
コード
https://github.com/hojonathanho/diffusion
解決した課題/先行研究との比較
技術・手法のポイント
評価指標
残された課題・議論
重要な引用
関連論文
参考
The text was updated successfully, but these errors were encountered: