Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Gradient Matching Generative Networks for 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 gradient matching loss to conditional WGAN training (ablation) | GMN · CUB, SUN, AWA | GZSL unseen-class accuracy (u): 55.9 -> 57.9; 53.6 -> 55.2; 55.2 -> 63.2 | Helped |
| add gradient matching loss to conditional WGAN training (ablation) | GMN · CUB, SUN, AWA | GZSL seen-class accuracy (s): 71.1 -> 71.2; 41.1 -> 40.8; 79.1 -> 78.8 | No clear effect |
| add gradient matching loss to conditional WGAN training (ablation) | GMN · CUB, SUN, AWA | ZSL T-1: 60.6 -> 64.6; 62.6 -> 64.1; 72.0 -> 73.9 | Helped |
| use an unconditional discriminator over seen and unseen data with gradient matching (ablation) | GMN · CUB, SUN, AWA | GZSL harmonic mean: 46.9 -> 47.5; 46.9 -> 65.0; 70.1 -> 74.8 | Mixed |
| train with a conditional discriminator instead of an unconditional seen-class discriminator, both with gradient matching (ablation) | GMN · CUB, SUN, AWA | GZSL harmonic mean: 46.5 -> 46.9; 62.4 -> 46.9; 62.7 -> 70.1 | Mixed |
| add synthesized examples for unseen classes to real seen-class training data (ablation) | supervised ZSL classifier · CUB, SUN, AWA | GZSL harmonic mean: 28.4 -> 46.9; 22.9 -> 46.9; 22.9 -> 70.1 | Helped |
| train a generator with the unconditional WGAN loss alone, without class supervision (ablation) | GMN generator · ZSL benchmarks | classifier performance | Hurt |
| increase the number of synthesized features per class (ablation) | feature-generation-based ZSL classifier · CUB, SUN, AWA | GZSL unseen-class accuracy (u) | Helped |
| increase the number of synthesized features per class (ablation) | feature-generation-based ZSL classifier · CUB, SUN, AWA | GZSL seen-class accuracy (s) | Hurt |
Closest papers by meaning
- Meta-Learning for Generalized Zero-Shot Learning (AAAI 2020 · similarity 0.92)
- Feature Generating Networks for Zero-Shot Learning (CVPR 2018 · similarity 0.91)
- Data-Free Generalized Zero-Shot Learning (AAAI 2024 · similarity 0.90)
- GenZSL: Generative Zero-Shot Learning Via Inductive Variational Autoencoder (ICML 2025 · similarity 0.90)
- Data Distribution Distilled Generative Model for Generalized Zero-Shot Recognition (AAAI 2024 · similarity 0.90)
- Improving Generalized Zero-Shot Learning by Exploring the Diverse Semantics from External Class Names (CVPR 2024 · similarity 0.88)
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.