AI Paper Atlas

Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning

Is Meta-Learning Out? Rethinking Unsupervised Few-Shot Classification with Limited Entropy

Yunchuan Guan, Yu Liu, Ke Zhou, Zhiqi Shen, Jenq-Neng Hwang, Serge Belongie, Lei Li

ICCV 2025 · Label-efficient and robust learning › Zero-/few-shot and meta-learning · approach: Meta-learning analysis and benchmarks · 5 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 DBSCAN heterogeneous task construction instead of K-means (ablation)MINO · Omniglot 5-way 1-shotaccuracy (%): 87.12 -> 93.81Improved
add DBSCAN heterogeneous task construction instead of K-means (ablation)MINO · CIFAR-100 zero-shotaccuracy (%): 37.58 -> 42.34Improved
use meta-learning instead of whole-class training (ablation)MINO · Omniglot 5-way 1-shotaccuracy (%): 74.32 -> 93.81Improved
use meta-learning instead of whole-class training (ablation)MINO · CIFAR-100 zero-shotaccuracy (%): 32.37 -> 42.34Improved
add stability-based meta-scaler (ablation)MINO · Omniglot 5-way 1-shotaccuracy (%): 91.56 -> 93.81Improved
add stability-based meta-scaler (ablation)MINO · STL-10 zero-shotaccuracy (%): 56.84 -> 58.74Improved
use bi-level meta-learning (MAML) instead of single-level WCT under 30% label noise (ablation)MAML · Omniglot 5-way 1-shotaccuracy (%): 64.65 -> 80.72Improved
use bi-level meta-learning (MAML) instead of single-level WCT under 30% label noise (ablation)MAML · Mini-ImageNet 5-way 1-shotaccuracy (%): 29.68 -> 37.51Improved
add dynamic head for heterogeneous tasks versus static head (ablation)DHM · Mini-ImageNet 5-10 way heterogeneousaccuracy (%): 41.63 -> 44.09Improved
use MAML instead of WCT in entropy-limited 3D few-shot (ablation)MAML · ShapeNetCore 5-way 1-shotaccuracy (%): 49.62 -> 54.03Improved
use MAML-style instead of ANIL-style inner updates (ablation)MINO · Omniglot 5-way 1-shottraining time to accuracy: 93.5 reached only by MAML (25717 s); ANIL faster to 90.0 (13987 vs 14723)Mixed
add DBSCAN instead of K-means for pseudo-labels (ablation)MINO · ImageNet zero-shotaccuracy (%): 29.89 -> 31.12Improved

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