AI Paper Atlas

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

Fine-Grained Zero-Shot Learning with DNA as Side Information

Sarkhan Badirli, Zeynep Akata, George Mohler, Christine Picard, Murat Dundar

NeurIPS 2021 · Label-efficient and robust learning › Zero-/few-shot and meta-learning · approach: Robust fine-tuning of zero-shot models · 19 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 BZSL instead of the next-best method (LsrGan) with DNA side informationBZSL · INSECTharmonic mean H: 17.75 -> 26.99Improved
replace word vectors with DNA as side informationBZSL · CUB GZSLharmonic mean H: 29.94 -> 34.97Improved
replace word vectors with DNA as side informationCADA-VAE · CUB GZSLharmonic mean H: 32.45 -> 25.48Hurt
replace word vectors with DNA as side informationCRNet · CUB GZSLharmonic mean H: 30.43 -> 15.93Hurt
use BZSL vs LsrGan with visual attributesBZSL · CUB GZSL visual attributesharmonic mean H: 51.89 -> 38.82Hurt
increase number of seen classes (unseen fixed at 10%) (ablation)BZSL · INSECTunseen accuracyImproved
increase number of seen classes (ablation)CADA-VAE · INSECTseen/unseen accuracyMixed
move kappa_1 away from [0.1, 1] range (much smaller or larger) (ablation)BZSL · INSECTunseen accuracyHurt

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