arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2003.01599cs.LGstat.ML
收藏

VQ-DRAW:一种序列离散VAE

VQ-DRAW: A Sequential Discrete VAE

Alex Nichol

摘要:

在本文中,我提出了VQ-DRAW,学习紧凑的离散表示的数据的算法。VQ-DRAW利用矢量量化效应使DRAW的顺序生成方案适应离散潜变量。我证明了VQ-DRAW可以有效地学习压缩来自各种常见数据集的图像,以及从这些数据集生成真实的样本,而无需自回归先验的帮助。

英文摘要:

In this paper, I present VQ-DRAW, an algorithm for learning compact discrete representations of data. VQ-DRAW leverages a vector quantization effect to adapt the sequential generation scheme of DRAW to discrete latent variables. I show that VQ-DRAW can effectively learn to compress images from a variety of common datasets, as well as generate realistic samples from these datasets with no help from an autoregressive prior.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑