AI Paper Atlas

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

Learning Discriminative Latent Attributes for Zero-Shot Classification

Huajie Jiang, Ruiping Wang, Shiguang Shan, Yi Yang, Xilin Chen

ICCV 2017 · Label-efficient and robust learning › Zero-/few-shot and meta-learning · approach: Generative feature synthesis for zero-shot · 112 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 the discriminative seen-class classifier constraint to the latent attribute dictionary (LA- to LA) (ablation)LAD · aP&YaccuracyImproved
learn latent attributes (linear combinations of attributes) instead of the attribute dictionary alone or the similarity space alone (ablation)LAD · aP&YaccuracyImproved
combine latent attribute space with similarity space (LA to LAS) (ablation)LAD · aP&Ymulti-class accuracy (%): 51.13 -> 53.74Improved
combine latent attribute space with similarity space (LA to LAS) (ablation)LAD · AwAmulti-class accuracy (%): 80.49 -> 82.48Improved
combine latent attribute space with similarity space (LA to LAS) (ablation)LAD · CUB-200multi-class accuracy (%): 56.36 -> 56.63Improved
combine latent attribute space with similarity space (LA to LAS) (ablation)LAD · SUN-Amulti-class accuracy (%): 83.50 -> 85.00Improved
use latent attribute dictionary versus SSE-ReLU on fine-grained CUB-200LAD (LAS) · CUB-200multi-class accuracy (%): 30.41 -> 56.63Improved
use LAD (LAS) versus JSLA, which models user-defined and discriminative attributes separatelyLAD (LAS) · AwAmulti-class accuracy (%): 82.81 -> 82.48Hurt
use LAD (LAS) versus Bucher et al. metric-learning ZSLLAD (LAS) · aP&Ymulti-class accuracy (%): 53.15 -> 53.74Improved

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