AI Paper Atlas

Research map / Probabilistic and scientific modeling / Variational and generative sampling

Cauchy–Schwarz Regularized Autoencoder

Linh Tran, Maja Pantic, Marc Peter Deisenroth

JMLR 2022 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Regularizing VAE latent information · 10 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace KL regularizer with Cauchy-Schwarz regularizerVAE (z=40) · Dynamic MNISTtest FID: 10.99 -> 7.88Improved
replace KL regularizer with Cauchy-Schwarz regularizerVAE (z=40) · Static MNISTtest FID: 10.04 -> 7.60Improved
replace KL regularizer with Cauchy-Schwarz regularizerVAE (z=40) · CIFAR10test FID: 11.36 -> 10.94Improved
add mixture-of-Gaussians prior (100 components) to CSRAECSRAE · Caltech101test FID: 17.15 -> 16.33Improved
add mixture prior (100 components) to CSRAECSRAE · Static MNISTtest FID: 7.60 -> 7.68No clear effect
add mixture prior (100 components) to CSRAECSRAE · CIFAR10test FID: 10.94 -> 10.65Improved
use MixtureCSRAE instead of VampPriorVAE for kNN on latent codesMixtureCSRAE (z=20) · Dynamic MNIST (k=3)kNN error (%): 3.98 -> 2.98Improved
use MixtureCSRAE instead of VampPriorVAE for kNN on latent codesMixtureCSRAE (z=20) · Caltech101 (k=10)kNN error (%): 38.52 -> 34.94Improved
use MixtureCSRAE instead of VampPriorVAE with 250 labelssemi-supervised VAE · Dynamic MNISTclassification error (%): 4.73 -> 4.12Improved
use MixtureCSRAE instead of VampPriorVAE with 4000 labelssemi-supervised VAE · Dynamic MNISTclassification error (%): 3.15 -> 2.75Improved
use MixtureCSRAE instead of VampPriorVAE on FERA2015 10% labelssemi-supervised VAE · FERA2015average F1: 23.07 -> 30.56Improved
increase labels from 10% to 25% for MixtureCSRAE on FERA2015MixtureCSRAE · FERA2015average F1: 30.56 -> 27.11Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy