AI Paper Atlas

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

Connectivity-Optimized Representation Learning via Persistent Homology

Christoph Hofer, Roland Kwitt, Marc Niethammer, Mandar Dixit

ICML 2019 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Disentangled and discrete latent models · 30 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
Use persistent-homology connectivity regularization in the autoencoder representationOne-class learning with 120 training samples per class · CIFAR-10AUC: 0.69 -> 0.76Improved
Use persistent-homology connectivity regularization in the autoencoder representationOne-class learning with 120 training samples per class · CIFAR-20AUC: 0.66 -> 0.72Improved
Use a connectivity-regularized representation rather than ADTOne-class learning · CIFAR-100, 120 samples/classAUC: 0.75 -> 0.79Improved
Transfer the autoencoder representation trained on CIFAR-100One-class learning · Tiny-ImageNetAUC: 0.73 vs. 0.72No clear effect
Transfer the autoencoder representation trained on CIFAR-100One-class learning · ImageNetAUC: 0.72 vs. 0.72No clear effect
Add 16 separate branches to control connectivity in low-dimensional subspaces (ablation)Connectivity-regularized autoencoder · CIFAR-10 one-class learningAUC: 0.70 -> 0.78Improved
Use 32 branches instead of no branching (ablation)Connectivity-regularized autoencoder · CIFAR-10 one-class learningAUC: 0.70 -> 0.75Improved
Reduce connectivity-loss weight to lambda=1.0 (ablation)16-branch autoencoder · CIFAR-100 training; CIFAR-10 one-class evaluationAUCHurt
Use connectivity-loss weight lambda>=10.0 rather than lambda=1.0 (ablation)16-branch autoencoder · CIFAR-100 trainingconnectivity lossImproved
Train with smaller batches, e.g. batch size 5 (ablation)Connectivity-loss toy mapping · Three-Gaussian toy datarange of death timesImproved
Increase latent dimensionality while keeping batch size fixed (ablation)Connectivity-loss toy mapping · Three-Gaussian toy datarange of death timesImproved

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