AI Paper Atlas

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

TransZero++: Cross Attribute-Guided Transformer for Zero-Shot Learning

Shiming Chen, Ziming Hong, Wenjin Hou, Guo-Sen Xie, Yibing Song, Jian Cun Zhao, Xinge You, Shuicheng Yan, Ling Shao

TPAMI 2022 · Label-efficient and robust learning › Zero-/few-shot and meta-learning · approach: Generative feature synthesis for zero-shot · 88 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 AVT sub-net (ablation)TransZero++ · CUBaccuracy: 49.0 -> 78.3Improved
add VAT sub-net (ablation)TransZero++ · CUBaccuracy: 75.6 -> 78.3Improved
add feature augmentation encoder (FAE) (ablation)TransZero++ · CUBaccuracy: 70.7 -> 78.3Improved
add feature augmentation (FA) in FAE (ablation)TransZero++ · CUBaccuracy: 76.4 -> 78.3Improved
add visual-semantic decoder (DEC) (ablation)TransZero++ · CUBaccuracy: 62.7 -> 78.3Improved
add self-calibration loss L_SC (ablation)TransZero++ · CUBH: 58.9 -> 70.4Improved
add attribute regression loss L_AR (ablation)TransZero++ · SUNaccuracy: 64.7 -> 67.6Improved
add semantical collaborative loss to AVT (ablation)TransZero++(AVT) · CUBaccuracy: 75.6 -> 77.6Improved
add semantical collaborative loss to VAT (ablation)TransZero++(VAT) · SUNaccuracy: 57.7 -> 63.3Improved
l2 distance instead of l1 in collaborative loss (ablation)TransZero++ · CUBaccuracy: 53.6 -> 78.3Improved
l2 instead of KL(p1||p2) in collaborative loss (ablation)TransZero++ · CUBaccuracy: 76.4 -> 78.3Improved
add TransZero++ features to f-VAEGANf-VAEGAN · CUBaccuracy: 61.0 -> 71.2Improved

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