Faithful Autoencoder Interpolation by Shaping the Latent Space

08/04/2020
by   Alon Oring, et al.
7

One of the fascinating properties of deep learning is the ability of the network to reveal the underlying factors characterizing elements in datasets of different types. Autoencoders represent an effective approach for computing these factors. Autoencoders have been studied in the context of their ability to interpolate between data points by decoding mixed latent vectors. However, this interpolation often incorporates disrupting artifacts or produces unrealistic images during reconstruction. We argue that these incongruities are due to the manifold structure of the latent space where interpolated latent vectors deviate from the data manifold. In this paper, we propose a regularization technique that shapes the latent space following the manifold assumption while enforcing the manifold to be smooth and convex. This regularization enables faithful interpolation between data points and can be used as a general regularization as well for avoiding overfitting and constraining the model complexity.

READ FULL TEXT

page 1

page 2

page 7

page 8

research
06/30/2021

Interventional Assays for the Latent Space of Autoencoders

The encoders and decoders of autoencoders effectively project the input ...
research
10/13/2021

Revisiting Latent-Space Interpolation via a Quantitative Evaluation Framework

Latent-space interpolation is commonly used to demonstrate the generaliz...
research
09/15/2023

A Geometric Perspective on Autoencoders

This paper presents the geometric aspect of the autoencoder framework, w...
research
07/19/2018

Understanding and Improving Interpolation in Autoencoders via an Adversarial Regularizer

Autoencoders provide a powerful framework for learning compressed repres...
research
06/18/2021

Generating various airfoil shapes with required lift coefficient using conditional variational autoencoders

Multiple shapes must be obtained in the mechanical design process to sat...
research
07/20/2022

Structural Causal 3D Reconstruction

This paper considers the problem of unsupervised 3D object reconstructio...
research
07/28/2023

Defocus Blur Synthesis and Deblurring via Interpolation and Extrapolation in Latent Space

Though modern microscopes have an autofocusing system to ensure optimal ...

Please sign up or login with your details

Forgot password? Click here to reset