Research map / Continual learning / Class-incremental
Gram-Calibrated Anchoring for Class-Incremental Learning
Qi Zhu, Ziang Gan, Libao Zhang
Read the paperCheck your idea against 137,217 papers
AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes Gram-Calibrated Anchoring (GCA), which uses class prototypes extracted once from a frozen pre-trained ViT as anchors for class-incremental learning under continual low-rank adaptation.
Problem
Class-incremental learning with pre-trained Vision Transformers under continual low-rank adaptation, where prototypes recomputed from the adapted model drift and cause forgetting.
Approach
Replace prototypes recomputed from the adapted model with anchor prototypes extracted once from the frozen pre-trained backbone, calibrated with a Gram-based term; a stability analysis bounds decision-boundary shifts.
Surprising findings
- Frozen-backbone anchor prototypes are effective substitutes for prototypes recomputed from the adapted model under continual low-rank adaptation (per abstract).
Closest papers by meaning
- HyCal: A Training-Free Prototype Calibration Method for Cross-Discipline Few-Shot Class-Incremental Learning (CVPR 2026 · similarity 0.88)
- Navigating Semantic Drift in Task-Agnostic Class-Incremental Learning (ICML 2025 · similarity 0.88)
- Class-Incremental Learning With Strong Pre-Trained Models (CVPR 2022 · similarity 0.88)
- Advancing Analytic Class-Incremental Learning through Vision-Language Calibration (ICML 2026 · similarity 0.88)
- Hierarchical Semantic Tree Anchoring for CLIP-Based Class-Incremental Learning (NeurIPS 2026 · similarity 0.87)
- Advancing Cross-domain Discriminability in Continual Learning of Vision-Language Models (NeurIPS 2024 · 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.
