Metric: average top-1 classification accuracy (higher is better)
| # | Model↕ | average top-1 classification accuracy▼ | Augmentations | Paper | Date↕ | Code |
|---|---|---|---|---|---|---|
| 1 | ZeroDiff | 77.3 | No | Exploring Data Efficiency in Zero-Shot Learning ... | 2024-06-05 | - |
| 2 | SPOT (VAEGAN) | 66.04 | No | Synthetic Sample Selection for Generalized Zero-... | 2023-04-06 | - |
| 3 | ZSL_TF-VAEGAN | 66 | No | Latent Embedding Feedback and Discriminative Fea... | 2020-03-17 | Code |
| 4 | f-VAEGAN | 64.7 | No | f-VAEGAN-D2: A Feature Generating Framework for ... | 2019-03-25 | - |
| 5 | DUET (Ours) | 64.4 | No | DUET: Cross-modal Semantic Grounding for Contras... | 2022-07-04 | Code |
| 6 | LisGAN | 61.7 | No | Leveraging the Invariant Side of Generative Zero... | 2019-04-08 | Code |
| 7 | TCN | 61.5 | No | Transferable Contrastive Network for Generalized... | 2019-08-16 | - |
| 8 | f-CLSWGAN | 60.8 | No | Feature Generating Networks for Zero-Shot Learning | 2017-12-04 | Code |
| 9 | Cycle-WGAN | 59.9 | No | Multi-modal Cycle-consistent Generalized Zero-Sh... | 2018-08-01 | Code |