Research map / Continual learning / Online and replay-based
Lifelong Variational Autoencoder via Online Adversarial Expansion Strategy
Fei Ye, Adrian G. Bors
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 changed | Where | Measured | Effect |
|---|---|---|---|
| Use OAES-ELBO instead of Dynamic-ELBO-OCM | VAE-DEM · Split MNIST | log-likelihood: -115.89 -> -103.93 | Improved |
| Use OAES-ELBO instead of Dynamic-ELBO-OCM | VAE-DEM · Split Fashion | log-likelihood: -237.69 -> -231.10 | Improved |
| Use OAES-ELBO instead of Dynamic-ELBO-OCM | VAE-DEM · Split MNIST-Fashion | log-likelihood: -187.49 -> -171.62 | Improved |
| Use OAES-ELBO instead of Dynamic-ELBO-OCM | VAE-DEM · Cross-domain | log-likelihood: -177.29 -> -165.29 | Improved |
| Use OAES instead of Dynamic-OCM | continual classifier · Split MNIST | accuracy: 94.02 ± 0.23 -> 94.69 ± 0.18 | Improved |
| Use OAES instead of Dynamic-OCM | continual classifier · Split CIFAR10 | accuracy: 49.16 ± 1.52 -> 52.16 ± 0.25 | Improved |
| Use OAES instead of Dynamic-OCM | continual classifier · Split CIFAR100 | accuracy: 21.79 ± 0.68 -> 26.01 ± 1.02 | Improved |
| Use OAES instead of CNDPM | continual classifier · M-S | accuracy: 47.64 -> 55.35 | Improved |
| Use OAES instead of CNDPM | continual classifier · M-C | accuracy: 66.25 -> 72.56 | Improved |
| Use OAES instead of CNDPM | continual classifier · Split IM | accuracy: 27.98 -> 29.62 | Improved |
| Lower the expansion threshold β (ablation) | OAES · Split MNIST | negative log-likelihood | Improved |
| Lower the expansion threshold β (ablation) | OAES · Split MNIST | number of components | Mixed |
Closest papers by meaning
- Lifelong Generative Modelling Using Dynamic Expansion Graph Model (AAAI 2022 · similarity 0.91)
- Continual Unsupervised Generative Modeling (TPAMI 2025 · similarity 0.91)
- BooVAE: Boosting Approach for Continual Learning of VAE (NeurIPS 2021 · similarity 0.90)
- Continual Variational Autoencoder Learning via Online Cooperative Memorization (ECCV 2022 · similarity 0.88)
- Task-Free Continual Generation and Representation Learning via Dynamic Expansionable Memory Cluster (AAAI 2024 · similarity 0.86)
- Dynamic Expansion Diffusion Learning for Lifelong Generative Modelling (AAAI 2025 · similarity 0.85)
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.
