AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Robust Disentanglement of a Few Factors at a Time using rPU-VAE

Benjamin Estermann, Markus Marks, Mehmet Fatih Yanik

NeurIPS 2020 · Label-efficient and robust learning › Recognition and long-tail · approach: Contrastive representation theory · 1 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 MIG instead of DCI Disentanglement as PBT eval function (ablation)PBT-S-VAE · dspritesMIGImproved
Use MIG instead of DCI Disentanglement as PBT eval function (ablation)PBT-S-VAE · dspritesDCI DisentanglementNo clear effect
Use MIG instead of DCI as PBT eval function (ablation)PBT-S-VAE · dspritesscore varianceImproved
Add PBT with 1000 labels versus Locatello et al. semi-supervised and model selection1000-PBT-S-VAE (MIG) · dspritesMIG and DCI DisentanglementImproved
Use all labels instead of 1000 labels in PBT-S-VAE (ablation)PBT-S-VAE (MIG) · dspritesMIG and DCI DisentanglementImproved
Replace MIG eval with unsupervised UDR eval (PBT-U-VAE)PBT-VAE · dspritesMIG / DCI Disentanglement: MIG 0.058 to 0.343, DCI 0.165 to 0.451Hurt
Add recursive dataset reduction and surrogate labels (rPU-VAE) to PBT-U-VAEPBT-U-VAE (UDR) · dsprites and shapes3dMIG / DCI DisentanglementImproved
rPU-VAE versus beta-VAE, FactorVAE, AnnealedVAE, DIP-VAE I/II, beta-TCVAErPU-VAE · dsprites and shapes3dMIG / DCI varianceImproved
Increase number of leaf-runs from 0 to 1 (ablation)rPU-VAE · dsprites and shapes3dMIG / DCI DisentanglementImproved
Increase number of leaf-runs from 1 to 3 (ablation)rPU-VAE · dsprites and shapes3dMIG / DCI DisentanglementImproved
Increase leaf-runs (ablation)rPU-VAE · dsprites and shapes3dscore varianceImproved
Unsupervised rPU-VAE versus supervised PBT-S-VAE upper boundrPU-VAE · shapes3dMIGHurt

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