AI Paper Atlas

Research map / Continual learning / Class-incremental

PCR: Proxy-Based Contrastive Replay for Online Class-Incremental Continual Learning

CVPR 2023 · Continual learning › Class-incremental · approach: Class-balanced knowledge retention · 75 citations (OpenAlex)

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's title only (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
Replace full-class proxy comparisons with batch-selected proxy pairsPCR · Split CIFAR100, buffer 1000final accuracy rate: 23.1 ±0.8 -> 25.6 ±0.6Helped
Replace full-class proxy comparisons with batch-selected proxy pairsPCR · Split MiniImageNet, buffer 1000final accuracy rate: 20.3 ±1.3 -> 24.2 ±0.9Helped
Add selected anchor-to-proxy pairs, excluding duplicate pairs (ablation)ER · Split CIFAR100, buffer 1000final accuracy rate: 17.6 ±0.9 -> 25.5 ±0.6Helped
Keep duplicate selected anchor-to-proxy pairs (ablation)ER+A · Split CIFAR100, buffer 1000final accuracy rate: 25.5 ±0.6 -> 25.6 ±0.6Helped
Add SCR's original anchor-to-sample pairs alongside proxy pairs (ablation)PCR · Split CIFAR100, buffer 1000final accuracy rate: 25.6 ±0.6 -> 24.1 ±0.5Hurt

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