AI Paper Atlas

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

Attributes2Classname: A Discriminative Model for Attribute-Based Unsupervised Zero-Shot Learning

Berkan Demirel, Ramazan Gokberk Cinbis, Nazli Ikizler-Cinbis

ICCV 2017 · Label-efficient and robust learning › Zero-/few-shot and meta-learning · approach: Generative feature synthesis for zero-shot · 58 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
learn word transform with predicate matrix (PBT) (ablation)GloVe class embeddings · AwAnormalized accuracy: 10.2 -> 60.7Improved
learn word transform with predicate matrix (PBT) (ablation)GloVe class embeddings · aPaYnormalized accuracy: 16.0 -> 29.4Improved
learn word transform with images (IBT) instead of identity (ablation)GloVe class embeddings · AwAnormalized accuracy: 10.2 -> 69.9Improved
learn word transform with images (IBT) instead of identity (ablation)GloVe class embeddings · aPaYnormalized accuracy: 16.0 -> 38.2Improved
image-based instead of predicate-based training (ablation)word transform · AwAnormalized accuracy: 60.7 -> 69.9Improved
image-based instead of predicate-based training (ablation)word transform · aPaYnormalized accuracy: 29.4 -> 38.2Improved
add external class names and predicate matrix to training (text-only) (ablation)PBT · AwAnormalized accuracy: 60.7 -> 63.0Improved
train on external class names only, no AwA classes (ablation)PBT · AwAnormalized accuracy: 44.0Improved
use pre-trained GloVe vectors instead of uniform random vectors (ablation)PBT · AwAtop-1 accuracy: 28.6Improved
use pre-trained GloVe vectors instead of uniform random vectors (ablation)IBT · AwAtop-1 accuracy: 13.6Improved

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