AI Paper Atlas

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

Zero-Shot Learning—A Comprehensive Evaluation of the Good, the Bad and the Ugly

Yongqin Xian, Christoph H. Lampert, Bernt Schiele, Zeynep Akata

TPAMI 2018 · Label-efficient and robust learning › Zero-/few-shot and meta-learning · approach: Generative feature synthesis for zero-shot · 1,800 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
evaluate on the proposed split instead of the standard splitGFZSL · AWA2 ZSLtop-1 accuracy: 79.3 -> 63.8Hurt
evaluate on the proposed split instead of the standard splitSAE · AWA1 ZSLtop-1 accuracy: 80.6 -> 53.0Hurt
evaluate on the proposed split instead of the standard splitSYNC · AWA2 ZSLtop-1 accuracy: 71.2 -> 49.3Hurt
evaluate on the proposed split instead of the standard splitDAP · AWA1 ZSLtop-1 accuracy: 57.1 -> 44.1Hurt
evaluate on the proposed split instead of the standard splitALE · aPY ZSLtop-1 accuracy: 30.9 -> 39.7Improved
tune SSE hyperparameters on validation setsSSE · AWA1 ZSLtop-1 accuracy: 64.9 -> 76.3Hurt
tune SSE parameters on the test setSSE · AWA1 ZSLtop-1 accuracy: 73.8 -> 76.3Hurt
add novelty detection to CMTCMT · aPY GZSLharmonic mean: 2.8 -> 19.0Improved
train on AWA2 instead of AWA1 (test on same set)12 ZSL methods · AWA1/AWA2 cross-datasettop-1 accuracy: p-value=0.007Mixed
switch from attribute datasets to ImageNet 20K classesall ZSL methods · ImageNet ZSLtop-1 accuracy: 0.96Hurt
use word2vec instead of attributes for a generative modelGFZSL · ImageNet 2H ZSLtop-1 accuracy: 60.6 -> 1.45Hurt
test on most populated instead of least populated classesSYNC · ImageNet ZSL 500 classestop-1 accuracy: 5.83 -> 15.83Improved

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