Research map / Continual learning / Online and replay-based
Continual Variational Autoencoder Learning via Online Cooperative Memorization
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 |
|---|---|---|---|
| replace random memory removal with OCM selection | single VAE · Split MNIST (log-likelihood) | log-likelihood: -150.79 -> -132.07 | Improved |
| replace random memory with OCM | single VAE · Split MNIST-Fashion (log-likelihood) | log-likelihood: -247.46 -> -215.62 | Improved |
| add dynamic expansion to OCM | VAE-ELBO-OCM · Cross-domain (log-likelihood) | log-likelihood: -201.31 -> -177.29 | Improved |
| add IWVAE bound (50 samples) to OCM on a single dataset | VAE-ELBO-OCM · Split MNIST (log-likelihood) | log-likelihood: -132.07 -> -127.11 | Improved |
| add IWVAE bound (50 samples) to OCM on mixed datasets | VAE-ELBO-OCM · Split MNIST-Fashion (log-likelihood) | log-likelihood: -215.62 -> -224.34 | Hurt |
| add IWVAE bound (50 samples) to OCM on cross-domain | VAE-ELBO-OCM · Cross-domain (log-likelihood) | log-likelihood: -201.31 -> -204.35 | Hurt |
| add OCM to single VAE | VAE-ELBO-Random · Split CIFAR10 | FID: 116.26 -> 98.76 | Improved |
| add expansion to OCM | VAE-ELBO-OCM · Split CIFAR10 | FID: 98.76 -> 92.99 | Improved |
| add OCM to single VAE on large-scale data | VAE-ELBO-Random · ImageNet 64x64 | FID: 145.36 -> 133.23 | Improved |
| use Dynamic-OCM instead of CoPE | CoPE · Split CIFAR10 classification | accuracy: 48.92 -> 49.16 | Improved |
| use Dynamic-OCM instead of CNDPM | CNDPM · Split CIFAR100 classification | accuracy: 20.10 -> 21.79 | Improved |
| single OCM model instead of expansion models LIMix/CNDPM | LIMix/CNDPM · Cross-domain (log-likelihood) | log-likelihood: -218.15 -> -201.31 | Improved |
Closest papers by meaning
- Continual Unsupervised Generative Modeling (TPAMI 2025 · similarity 0.91)
- Task-Free Continual Generation and Representation Learning via Dynamic Expansionable Memory Cluster (AAAI 2024 · similarity 0.89)
- Forgetting, Ignorance or Myopia: Revisiting Key Challenges in Online Continual Learning (NeurIPS 2024 · similarity 0.87)
- Online Task-Free Continual Learning via Dynamic Expansionable Memory Distribution (CVPR 2025 · similarity 0.87)
- Temporal-Difference Variational Continual Learning (NeurIPS 2025 · similarity 0.86)
- Continual Unsupervised Generative Modelling via Online Optimal Transport (AAAI 2025 · similarity 0.86)
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.
