paper-with-me

홈 › Papers

Regularized Autoencoders via Relaxed Injective Probability Flow

2020-02-20 · Abhishek Kumar, Ben Poole, Kevin Murphy

Invertible flow-based generative models are an effective method for learning to generate samples, while allowing for tractable likelihood computation and inference. However, the invertibility requirement restricts models to have the same latent dimensionality as the inputs. This imposes significant architectural, memory, and computational costs, making them more challenging to scale than other classes of generative models such as Variational Autoencoders (VAEs). We propose a generative model based on probability flows that does away with the bijectivity requirement on the model and only assumes injectivity. This also provides another perspective on regularized autoencoders (RAEs), with our final objectives resembling RAEs with specific regularizers that are derived by lower bounding the probability flow objective. We empirically demonstrate the promise of the proposed model, improving over VAEs and AEs in terms of sample quality.

📄 PDF Abstract BibTeX arXiv:2002.08927

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Nonlinear Isometric Manifold Learning for Injective Normalizing Flows

2022-03-08 · Eike Cramer, Felix Rauh, Alexander Mitsos, Raúl Tempone 외

To model manifold data using normalizing flows, we employ isometric autoencoders to design embeddings with explicit inverses that do not distort the probability distribution. Using isometries separates manifold learning …

Density EstimationModel Selection

Universal Joint Approximation of Manifolds and Densities by Simple Injective Flows

2021-10-08 · Michael Puthawala, Matti Lassas, Ivan Dokmanić, Maarten de Hoop

We study approximation of probability measures supported on $n$-dimensional manifolds embedded in $\mathbb{R}^m$ by injective flows -- neural networks composed of invertible flows and injective layers. We show that in ge…

Bi-Lipschitz Autoencoder With Injectivity Guarantee

2026-04-08 · Qipeng Zhan, Zhuoping Zhou, Zexuan Wang, Qi Long 외 arxiv

Autoencoders are widely used for dimensionality reduction, based on the assumption that high-dimensional data lies on low-dimensional manifolds. Regularized autoencoders aim to preserve manifold geometry during dimension…

Dimensionality Reduction

Principal Manifold Flows

2022-02-14 · Edmond Cunningham, Adam Cobb, Susmit Jha

Normalizing flows map an independent set of latent variables to their samples using a bijective transformation. Despite the exact correspondence between samples and latent variables, their high level relationship is not …

Density Estimation

Injective flows for star-like manifolds

2024-06-13 · Marcello Massimo Negri, Jonathan Aellen, Volker Roth

Normalizing Flows (NFs) are powerful and efficient models for density estimation. When modeling densities on manifolds, NFs can be generalized to injective flows but the Jacobian determinant becomes computationally prohi…

Density EstimationVariational Inference