AI Paper Atlas

Research map / Continual learning / Online and replay-based

Wasserstein Expansible Variational Autoencoder for Discriminative and Generative Continual Learning

Fei Ye, Adrian G. Bors

ICCV 2023 · Continual learning › Online and replay-based · approach: Unified and unsupervised continual frameworks · 4 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
add energy-based sample selection (ablation)WEVAE · Split MNIST (density estimation)log-likelihood: -99.29 -> -89.66Improved
add energy-based sample selection (ablation)WEVAE · Split CIFAR10 (classification)accuracy (%): 51.38 -> 55.23Improved
add energy-based sample selection (ablation)WEVAE · Split CIFAR100 (classification)accuracy (%): 23.18 -> 25.07Improved
add energy-based sample selection (ablation)WEVAE · M-C (MNIST+CIFAR10 stream)accuracy (%): 56.98 -> 59.07Improved
use Wasserstein-based expansion instead of prior dynamic expansion (without sample selection)WEVAE-NoS vs Dynamic-OCM · Split CIFAR10 (classification)accuracy (%): 49.16 -> 51.38Improved
lower the expansion threshold lambda (more components) (ablation)WEVAE · Split MNIST (density estimation)negative log-likelihoodNo clear effect
vary memory buffer size (ablation)WEVAE vs baselines · Cross-Domainnegative log-likelihoodImproved
change batch size (ablation)WEVAE · Split MNISTnegative log-likelihoodNo clear effect

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