AI Paper Atlas

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

Few-Shot Image Classification: Just Use a Library of Pre-Trained Feature Extractors and a Simple Classifier

Arkabandhu Chowdhury, Mingchao Jiang, Swarat Chaudhuri, Chris Jermaine

ICCV 2021 · Label-efficient and robust learning › Zero-/few-shot and meta-learning · approach: Embedding adaptation for few-shot classification · 38 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
concatenate features of all nine library CNNs instead of using the best single CNN (ablation)MLP few-shot learner · Aircraft 20-way 5-shotaccuracy: 44.6 -> 49.5Improved
concatenate features of all nine library CNNs instead of using the best single CNN (ablation)MLP few-shot learner · Aircraft 40-way 5-shotaccuracy: 35.9 -> 41.2Improved
concatenate features of all nine library CNNs instead of using the best single CNN (ablation)MLP few-shot learner · Aircraft 5-way 5-shotaccuracy: 66.0 -> 68.9Improved
use one MLP on concatenated features instead of a hard ensemble of per-CNN classifiers (ablation)Full Library · Aircraft 20-way 5-shotaccuracy: 46.6 -> 49.5Improved
use one MLP on concatenated features instead of a soft ensemble of per-CNN classifiers (ablation)Full Library · Aircraft 40-way 5-shotaccuracy: 39.0 -> 41.2Improved
use concatenated features instead of ensembling on 5-way tasks (ablation)Full Library · FC100 5-way 5-shotaccuracy: 79.1 vs 80.5No clear effect
use library of nine ILSVRC2012 CNNs instead of a single BiT CNN trained on full ImageNet (ablation)Full Library · Aircraft 40-way 5-shotaccuracy: 33.0 -> 41.2Improved
use library of nine ILSVRC2012 CNNs instead of a single BiT CNN trained on full ImageNet (ablation)Full Library · VGG Flower 5-way 5-shotaccuracy: 97.4 vs 99.7Hurt
use library of nine ILSVRC2012 CNNs instead of a single BiT CNN (ablation)Full Library · Traffic Sign 20-way 5-shotaccuracy: 53.5 -> 70.4Improved
replace meta-learning/transfer baselines with a library-based learnerbest library-based learner · Aircraft 5-way 5-shotaccuracy: 47.6 -> 66.0Improved
replace SUR with a library-based learnerbest library-based learner · Omniglot 5-way 5-shotaccuracy: 98.7 vs 96.7Hurt
train on 1-shot vs full data without regularization: compare learned weightslinear classifier on 13,984 features · Traffic Sign 40-wayPearson correlation: 0.95 on Aircraft vs 0.18 on Traffic SignMixed

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