AI Paper Atlas

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

A Broader Study of Cross-Domain Few-Shot Learning

Yunhui Guo, Noel C. Codella, Leonid Karlinsky, James V. Codella, John R. Smith, Kate Saenko, Tajana Rosing, Rogerio Feris

ECCV 2020 · Label-efficient and robust learning › Zero-/few-shot and meta-learning · approach: Fine-grained and cross-domain few-shot baselines · 374 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
use ProtoNet instead of MetaOpt (older vs newer meta-learner)meta-learning on BSCD-FSL · BSCD-FSL averageaccuracy: 59.78 vs 57.35Improved
add feature-wise transform (FWT)ProtoNet · BSCD-FSL averageaccuracy: 59.78 -> 56.72Hurt
add feature-wise transform (FWT)MatchingNet · BSCD-FSL averageaccuracy: 50.21 -> 46.55Hurt
add feature-wise transform (FWT)RelationNet · BSCD-FSL averageaccuracy: 54.48 -> 52.6Hurt
replace meta-learning with simple fine-tuningProtoNet · BSCD-FSL averageaccuracy: 59.78 -> 67.23Improved
use MatchingNet instead of random embeddingfixed feature extractor · BSCD-FSL averageaccuracy: 53.99 vs 50.21Hurt
add transductive batch statistics to fine-tuningfine-tune all layers · BSCD-FSL averageaccuracy: 67.23 -> 68.14Improved
use pretrained ImageNet-like weights instead of random weightsfixed feature extractor · BSCD-FSL averageaccuracy: 53.99 -> 64.24Improved
use mean-centroid classifier instead of linear classifier (ablation)fine-tuning · ISIC 5-way 50-shotaccuracy: 66.48 -> 61.57Hurt
use cosine-similarity classifier instead of linear classifier (ablation)fine-tuning · EuroSAT 5-way 50-shotaccuracy: 91.34 -> 88.83Hurt
use cosine-similarity classifier instead of linear classifier (ablation)fine-tuning · ChestX 5-way 5-shotaccuracy: 25.97 -> 26.95Improved
use incremental multi-model selection instead of concatenating all embeddings (ablation)multiple pretrained models · BSCD-FSL averageaccuracy: 68.22 -> 68.69Improved

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