AI Paper Atlas

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

Class-Aware Patch Embedding Adaptation for Few-Shot Image Classification

Fusheng Hao, Fengxiang He, Liu Liu, Fuxiang Wu, Dacheng Tao, Jun Cheng

ICCV 2023 · Label-efficient and robust learning › Zero-/few-shot and meta-learning · approach: Fine-grained and cross-domain few-shot baselines · 71 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 class-aware adaptation of patch embeddings (lambda 0 to 2) (ablation)CPEA · miniImageNet 1-shotaccuracy: 70.40 -> 71.97Improved
add class-aware adaptation of patch embeddings (lambda 0 to 2) (ablation)CPEA · miniImageNet 5-shotaccuracy: 85.05 -> 87.06Improved
increase class-aware factor beyond 2 (ablation)CPEA · miniImageNetaccuracy: 87.06 -> 87.22 (5-shot lambda 8)No clear effect
add projection head (ablation)CPEA · miniImageNet 1-shotaccuracy: 70.74 -> 71.99Improved
square the cosine similarity in the dense score matrix (ablation)CPEA · miniImageNet 1-shotaccuracy: 71.09 -> 71.97Improved
use MIM pretraining instead of supervised DeiT (ablation)CPEA · miniImageNet 1-shotaccuracy: 28.58 -> 71.97Improved
use MIM pretraining instead of DINO (ablation)CPEA · miniImageNet 1-shotaccuracy: 70.65 -> 71.97Improved
use more than one class-agnostic embedding (ablation)CPEA · miniImageNet 1-shotaccuracy: 71.97 -> 70.57 (4 embeddings)Hurt
replace online-optimized FewTURE classifier with CPEA dense similarity (ablation)ViT-S/16 backbone · miniImageNet 5-shotaccuracy: 84.51 -> 87.06Improved
replace FewTURE online optimization with CPEAinference · miniImageNet 5-shotinference time (ms): 168.6 -> 1.352Improved
increase input resolution from 224 to 448 (ablation)CPEA · miniImageNet 5-shotaccuracy: 87.06 -> 88.00Improved
replace cosine similarity with scaled Euclidean metrics (ablation)CPEA · miniImageNet 5-shotaccuracy: 87.06 -> 85.78 (best Euclidean variant)Hurt

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