Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Distinguishing Unseen From Seen for Generalized Zero-Shot Learning
Check 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 |
|---|---|---|---|
| Add Wasserstein distance minimization to align visual and semantic latent representations (ablation) | DUS-VAE · AwA1, AwA2, CUB, SUN, and FLO | H: 54.63 -> 79.72; 49.71 -> 77.56; 66.74 -> 70.52; 69.76 -> 73.10; 75.25 -> 76.46 | Helped |
| Add fictitious classes generated using unseen semantic attributes (ablation) | DUS-VAE · AwA1, AwA2, CUB, SUN, and FLO | H: 71.15 -> 79.72; 71.00 -> 77.56; 69.44 -> 70.52; 71.06 -> 73.10; 76.07 -> 76.46 | Helped |
| Add cycle-consistent reconstruction between visual and semantic modalities (ablation) | DUS-VAE · AwA1, AwA2, CUB, SUN, and FLO | H: 78.26 -> 79.72; 75.64 -> 77.56; 70.39 -> 70.52; 73.00 -> 73.10; 76.14 -> 76.46 | Mixed |
| Add classification loss (ablation) | DUS-VAE · AwA1, AwA2, CUB, SUN, and FLO | AUC: 99.49 -> 99.90; 98.99 -> 99.30; 99.99 -> 99.99; 99.88 -> 99.98; 99.93 -> 99.97 | Mixed |
| Use the proposed seen/unseen domain separation with the unseen expert instead of OOD | GZSL · AwA1 | H: 72.6 -> 82.9 | Helped |
| Use the proposed seen/unseen domain separation with the unseen expert instead of OOD | GZSL · AwA2 | H: 70.3 -> 80.1 | Helped |
| Use the proposed seen/unseen domain separation with the unseen expert instead of GEM-ZSL | GZSL · CUB | H: 70.4 -> 70.3 | No clear effect |
| Use the proposed seen/unseen domain separation with the unseen expert instead of OOD | GZSL · SUN | H: 71.9 -> 75.2 | Helped |
| Use the proposed seen/unseen domain separation with the unseen expert instead of FREE | GZSL · FLO | H: 75.0 -> 74.3 | Hurt |
Closest papers by meaning
- Improving Generalized Zero-Shot Learning by Exploring the Diverse Semantics from External Class Names (CVPR 2024 · similarity 0.94)
- Adaptive and Generative Zero-Shot Learning (ICLR 2021 · similarity 0.93)
- Generalized Zero-Shot Recognition Based on Visually Semantic Embedding (CVPR 2019 · similarity 0.92)
- A Variational Autoencoder with Deep Embedding Model for Generalized Zero-Shot Learning (AAAI 2020 · similarity 0.92)
- ZeroDiff: Solidified Visual-semantic Correlation in Zero-Shot Learning (ICLR 2025 · similarity 0.90)
- Data Distribution Distilled Generative Model for Generalized Zero-Shot Recognition (AAAI 2024 · similarity 0.90)
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.