AI Paper Atlas

Research map / Label-efficient and robust learning / Self-supervised and contrastive

Few-Shot Classification with Contrastive Learning

Zhanyuan Yang, Jinghua Wang, Yingying Zhu

ECCV 2022 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Inductive bias and minimal representations

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 all pre-training contrastive losses to CE (ablation)FEAT baseline · miniImageNet 5-way 1-shotaccuracy: 66.58 -> 69.53Improved
add all pre-training contrastive losses to CE (ablation)FEAT baseline · miniImageNet 5-way 5-shotaccuracy: 81.92 -> 84.33Improved
add cross-view episodic training (ablation)L_CE+L_CL · miniImageNet 5-way 1-shotaccuracy: 69.53 -> 69.89Improved
add cross-view episodic training (ablation)L_CE+L_CL · CIFAR-FS 5-way 1-shotaccuracy: 76.15 -> 77.02Improved
add distance-scaled contrastive loss L_info (ablation)L_CE+L_CL · tieredImageNet 5-way 1-shotaccuracy: 71.83 -> 72.51Improved
add both CVET and L_info (ablation)L_CE+L_CL · miniImageNet 5-way 1-shotaccuracy: 69.53 -> 70.19Improved
add local self-supervised loss only (ablation)L_CE · miniImageNet 5-way 1-shotaccuracy: 66.58 -> 69.33Improved
add global supervised contrastive loss only (ablation)L_CE · miniImageNet 5-way 1-shotaccuracy: 66.58 -> 67.88Improved
add vector-map module (ablation)global contrastive losses · miniImageNet 5-way 1-shotaccuracy: 68.68 -> 69.29Improved
add map-map module (ablation)global contrastive losses · miniImageNet 5-way 5-shotaccuracy: 84.17 -> 84.25Improved
use SimCLR-style instead of standard augmentation (ablation)Ours · miniImageNet 5-way 1-shotaccuracy: 67.84 -> 70.19Improved
use AutoAugment instead of SimCLR-style augmentation (ablation)Ours · miniImageNet 5-way 1-shotaccuracy: 70.19 -> 68.45Hurt

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