Bolsa Institucional De Valores S A Institucional De Valores, Bolsa Institucional De Valores
Please join the Simons Foundation and our generous member organizations in supporting travelhome.vn during our giving campaign September 23-27. 100% of your contribution will fund improvements and new initiatives to benefit travelhome.vn”s global scientific community.
Đang xem: Bolsa institucional de valores s
All fields Title Author Abstract Comments Journal reference ACM classification MSC classification Report number travelhome.vn identifier DOI ORCID travelhome.vn author ID Help pages Full text
Xem thêm: {Review} Sữa Rửa Mặt St Ives Trà Xanh Review ) Sữa Rửa Mặt St
Download PDF Abstract: With the introduction of the variational autoencoder (VAE), probabilisticlatent variable models have received renewed attention as powerful generativemodels. However, their performance in terms of test likelihood and quality ofgenerated samples has been surpassed by autoregressive models withoutstochastic units. Furthermore, flow-based models have recently been shown to bean attractive alternative that scales well to high-dimensional data. In thispaper we close the performance gap by constructing VAE models that caneffectively utilize a deep hierarchy of stochastic variables and model complexcovariance structures. We introduce the Bidirectional-Inference VariationalAutoencoder (BIVA), characterized by a skip-connected generative model and aninference network formed by a bidirectional stochastic inference path. We showthat BIVA reaches state-of-the-art test likelihoods, generates sharp andcoherent natural images, and uses the hierarchy of latent variables to capturedifferent aspects of the data distribution. We observe that BIVA, in contrastto recent results, can be used for anomaly detection. We attribute this to thehierarchy of latent variables which is able to extract high-level semanticfeatures. Finally, we extend BIVA to semi-supervised classification tasks andshow that it performs comparably to state-of-the-art results by generativeadversarial networks.
| Subjects: | Machine Learning (stat.ML); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG) |
| Cite as: | travelhome.vn:1902.02102 |
| (or travelhome.vn:1902.02102v3 |
Xem thêm: Biệt Thự Của Ca Sĩ Đàm Vĩnh Hưng, Biệt Thự Khủng Của Đàm Vĩnh Hưng Tại Mỹ
Submission history
From: Lars Maaløe
Bình luận