AI Paper Atlas

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

Weak-shot Fine-grained Classification via Similarity Transfer

Junjie Chen, Li Niu, Liu Liu, Liqing Zhang

NeurIPS 2021 · Label-efficient and robust learning › Zero-/few-shot and meta-learning · approach: Embedding adaptation for few-shot classification · 9 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 SimTrans instead of the strongest listed baseline, Fusion2Weak-shot fine-grained classification · Caraccuracy: 86.9 -> 89.8Improved
Use SimTrans instead of the strongest listed baseline, Fusion2Weak-shot fine-grained classification · CUBaccuracy: 88.5 -> 91.7Improved
Use SimTrans instead of the strongest listed baseline, Fusion2Weak-shot fine-grained classification · FGVCaccuracy: 90.0 -> 92.8Improved
Add sample weighting to classification loss (ablation)CUB classifier · CUBaccuracy: 85.4 -> 90.3Improved
Add graph regularization to classification loss (ablation)CUB classifier · CUBaccuracy: 85.4 -> 89.5Improved
Add sample weighting and graph regularization (ablation)CUB classifier · CUBaccuracy: 85.4 -> 91.2Improved
Add adversarial loss to SimNet while using sample weighting (ablation)CUB classifier · CUBaccuracy: 90.3 -> 90.7Improved
Add adversarial loss to SimNet while using graph regularization (ablation)CUB classifier · CUBaccuracy: 89.5 -> 90.1Improved
Use the full method rather than classification loss alone (ablation)CUB classifier · CUBaccuracy: 85.4 -> 91.7Improved
Support the classifier with SimNet similarity learned from clean base categories rather than noisy novel data (ablation)CUB classifier · CUBaccuracy: 87.6 -> 91.7Improved
Use end-to-end SimNet similarity rather than Euclidean similarity from base data (ablation)CUB classifier · CUBaccuracy: 88.8 -> 91.7Improved
Reduce base training set to 50 categories with 5 images per category (ablation)SimTrans classifier · CUBaccuracy: 91.7 -> 89.3Hurt

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