Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
ZeroDiff++: Generative Test-Time Adaptation for Zero-shot Learning
Zihan Ye, Shreyank N Gowda, Kaile Du, Weijian Luo, Ling Shao
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add diffusion augmentation to the f-VAEGAN baseline (ablation) | ZeroDiff++ · AWA2 ZSL | top-1 accuracy: 76.0 -> 80.0 | Improved |
| add SC representations and mutual learning loss on top of diffusion augmentation (ablation) | ZeroDiff++ · AWA2 ZSL | top-1 accuracy: 80.0 -> 86.4 | Improved |
| add DiffTTA to the model with DiffAug, SC and L_mu (ablation) | ZeroDiff++ · AWA2 GZSL | harmonic mean H: 76.9 -> 78.7 | Improved |
| add DiffGen to the model with DiffAug, SC and L_mu (ablation) | ZeroDiff++ · AWA2 GZSL | harmonic mean H: 76.9 -> 84.4 | Improved |
| add DiffTTA when DiffGen is used (ablation) | ZeroDiff++ · AWA2 GZSL | harmonic mean H: 88.0 -> 92.3 | Improved |
| add DiffTTA when generating with fully-noised FNGen (ablation) | ZeroDiff++ · AWA2 GZSL | harmonic mean H: 73.8 -> 73.3 | No clear effect |
| replace FNGen inference with DiffGen (ablation) | ZeroDiff++ · AWA2 GZSL | harmonic mean H: 73.8 -> 88.0 | Improved |
| replace DiffusionZSL reconstruction-error inference with DiffGen (ablation) | ZeroDiff++ · AWA2 GZSL | harmonic mean H: 11.9 -> 88.0 | Improved |
| use real paired SC representation instead of generated SC in DiffGen (ablation) | DiffGen · AWA2 GZSL | harmonic mean H: 81.8 -> 91.9 | Improved |
| use intermediate diffusion time t=2 rather than extreme t in DiffGen with real SC (ablation) | DiffGen · AWA2 GZSL | harmonic mean H: 91.4 -> 91.9 | Improved |
| randomize diffusion time t in DiffGen instead of fixing it (ablation) | DiffGen · AWA2 GZSL | harmonic mean H: 91.9 -> 91.6 | No clear effect |
| add contrastive representation to the final classifier input (V+C) (ablation) | ZeroDiff++ · AWA2 GZSL | harmonic mean H: 90.6 -> 92.3 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- ZeroDiff — ZeroDiff: Solidified Visual-semantic Correlation in Zero-Shot Learning (ICLR 2025) · better on 9 of 9 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- ZeroDiff: Solidified Visual-semantic Correlation in Zero-Shot Learning (ICLR 2025 · similarity 0.96)
- GenZSL: Generative Zero-Shot Learning Via Inductive Variational Autoencoder (ICML 2025 · similarity 0.90)
- Improving Generalized Zero-Shot Learning by Exploring the Diverse Semantics from External Class Names (CVPR 2024 · similarity 0.90)
- Visual-Augmented Dynamic Semantic Prototype for Generative Zero-Shot Learning (CVPR 2024 · similarity 0.90)
- Data Distribution Distilled Generative Model for Generalized Zero-Shot Recognition (AAAI 2024 · similarity 0.90)
- ZeroMamba: Exploring Visual State Space Model for Zero-Shot Learning (AAAI 2025 · 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.
