Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Make an Omelette with Breaking Eggs: Zero-Shot Learning for Novel Attribute Synthesis
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| Use ZSLA instead of the strongest baseline | novel attribute detection on CUB with 32 seen attributes · CUB | mAUROC: 0.626 -> 0.689 | Helped |
| Use ZSLA instead of the strongest baseline | novel attribute detection on CUB with 32 seen attributes · CUB | mAP@50: 0.223 -> 0.320 | Helped |
| Use ZSLA instead of the strongest baseline | novel attribute detection on CUB with 32 seen attributes · CUB | mLA: 0.782 -> 0.846 | Helped |
| Re-annotate with ZSLA-synthesized attributes instead of using 32 manual attributes | CADAVAE GZSL training on CUB · CUB GZSL | H: 33.4 -> 53.2 | Helped |
| Re-annotate with ZSLA-synthesized attributes instead of using 32 manual attributes | TFVAEGAN GZSL training on CUB · CUB GZSL | H: 37.1 -> 57.4 | Helped |
| Re-annotate with ZSLA-synthesized attributes instead of using 32 manual attributes | ALE GZSL training on CUB · CUB GZSL | H: 13.7 -> 36.1 | Helped |
| Re-annotate with ZSLA-synthesized attributes instead of using 32 manual attributes | ESZSL GZSL training on CUB · CUB GZSL | H: 15.9 -> 26.2 | Helped |
| Add uni-modal constraint (ablation) | ZSLA novel attribute detectors without location information · CUB | mLA: .348 -> .613 | Helped |
| Add ground-truth location information (ablation) | ZSLA novel attribute detectors with UMC · CUB | mLA: .613 -> .846 | Helped |
| Add location information and UMC (ablation) | ZSLA novel attribute detectors · CUB | mAUROC: .702 -> .689 | No clear effect |
| Add location information and UMC (ablation) | ZSLA novel attribute detectors · CUB | mAP@50: .325 -> .320 | No clear effect |
Closest papers by meaning
- Zero-Shot Learning With Attribute Selection (AAAI 2018 · similarity 0.92)
- Zero-Shot Recognition via Direct Classifier Learning with Transferred Samples and Pseudo Labels (AAAI 2017 · similarity 0.90)
- Recovering the Missing Link: Predicting Class-Attribute Associations for Unsupervised Zero-Shot Learning (CVPR 2016 · similarity 0.89)
- Attribute Propagation Network for Graph Zero-Shot Learning (AAAI 2020 · similarity 0.89)
- Data-Free Generalized Zero-Shot Learning (AAAI 2024 · similarity 0.87)
- ZeroDiff: Solidified Visual-semantic Correlation in Zero-Shot Learning (ICLR 2025 · similarity 0.87)
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.