AI Paper Atlas

Research map / Continual learning / Class-incremental

AREA: Attribute Extraction and Aggregation for CLIP-Based Class-Incremental Learning

Zhen-Hao Xie, Yu-Cheng Shi, Da-Wei Zhou

ICML 2026 · Continual learning › Class-incremental · approach: Exemplar-efficient class-incremental learning · 0 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 PGA-anchored attributes with multi-modal aggregation to zero-shot CLIP (ablation)AREA · CIFAR100 B0 Inc10accuracyImproved
add VIB loss (intervention and compression terms) (ablation)AREA · CIFAR100 B0 Inc10accuracyImproved
add optimal transport routing at inference (ablation)AREA · CIFAR100 B0 Inc10accuracyImproved
use OT routing instead of cosine routing (ablation)AREA · CIFAR100 B0 Inc10, final stagefinal accuracy: +3.39% accuracy, +2.9 ms/sampleImproved
use PGA anchors instead of SVD anchors (ablation)AREA attribute extraction · ObjectNet attribute similaritysimilarity to fine-grained descriptionsImproved
caption only 5% to 20% of training data with MLLM (ablation)AREA · CIFAR100accuracy: 87.9% at 5%; drop under 1% from 100% to 20%No clear effect
swap captioning MLLM from GPT-5 to LLaVA-v1.6-34b (ablation)AREA · Aircraft, CIFAR, CUBaverage accuracy: 71.03 -> 70.89 (Aircraft); 89.24 -> 88.98 (CIFAR)No clear effect
vary lambda_int and lambda_comp in 0.5 to 1.5 (ablation)AREA · CIFAR100final average accuracy: range 87.86 to 89.24No clear effect
AREA versus MG-CLIP on CIFAR average accuracyCLIP-based CIL · CIFAR B0 Inc10average accuracy: 89.74 (MG-CLIP) vs 89.24Hurt
AREA versus RAPF on hard datasetsCLIP-based CIL · Aircraft B0 Inc10last accuracy: 23.61 -> 61.78Improved
add attribute anchors to prevent extraction drift (ablation)AREA · SUN exampleprediction rank of correct classImproved

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