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 1702.08658cs.LGstat.ML
收藏

深入理解变分自编码模型

Towards Deeper Understanding of Variational Autoencoding Models

Stanford University(斯坦福大学)

Shengjia Zhao, Jiaming Song, Stefano Ermon

摘要:

我们提出了一个新的家庭变分自动编码模型的优化标准,推广了标准的证据下限。我们提供的条件下,他们恢复的数据分布和学习潜在的功能,并正式显示,常见的问题,如模糊的样本和无信息的潜在功能时,这些条件不满足。基于这些新的见解,我们提出了一个新的顺序VAE模型,可以在LSUN图像数据集上生成基于像素级重建损失的清晰样本,并提出了一个优化标准,鼓励信息潜在特征的无监督学习。

英文摘要:

We propose a new family of optimization criteria for variational auto-encoding models, generalizing the standard evidence lower bound. We provide conditions under which they recover the data distribution and learn latent features, and formally show that common issues such as blurry samples and uninformative latent features arise when these conditions are not met. Based on these new insights, we propose a new sequential VAE model that can generate sharp samples on the LSUN image dataset based on pixel-wise reconstruction loss, and propose an optimization criterion that encourages unsupervised learning of informative latent features.

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。

↑