Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Towards Recognizing Unseen Categories in Unseen Domains
Massimiliano Mancini, Zeynep Akata, Elisa Ricci, Barbara Caputo
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 feature-level mixing (L_M-F) to the aggregate baseline (ablation) | AGG on PACS · PACS (ResNet-18) | average accuracy: 78.5 -> 81.1 | Improved |
| add image-level mixing (L_M-IMG) only (ablation) | AGG on PACS · PACS (ResNet-18) | average accuracy: 78.5 -> 76.3 | Hurt |
| add image-level mixing only, Sketch domain (ablation) | AGG on PACS · PACS Sketch | accuracy: 69.4 -> 59.5 | Hurt |
| add image-level mixing to feature-level mixing (ablation) | CuMix without curriculum · PACS (ResNet-18) | average accuracy: 81.1 -> 81.2 | Improved |
| add curriculum mixing strategy (ablation) | CuMix (image+feature mixing) · PACS (ResNet-18) | average accuracy: 81.2 -> 81.6 | Improved |
| add curriculum mixing strategy, Sketch domain (ablation) | CuMix · PACS Sketch | accuracy: 71.5 -> 72.6 | Improved |
| add mixing (CuMix) to the ZSL baseline on CUB (ablation) | cross-entropy baseline · CUB ZSL | accuracy: 52.4 -> 60.4 | Improved |
| add mixing (CuMix) on AWA1 (ablation) | cross-entropy baseline · AWA1 ZSL | accuracy: 62.5 -> 64.0 (GFZSL 68.3) | Mixed |
| add CuMix to SPNet for ZSL+DG | SPNet · DomainNet ZSL+DG | average per-class accuracy: 19.4 -> 20.7 | Improved |
| add standard mixup to SPNet | SPNet · DomainNet ZSL+DG | average per-class accuracy: 19.4 -> 19.7 | No clear effect |
| add DANN to SPNet | SPNet · DomainNet ZSL+DG | average per-class accuracy: 19.4 -> 19.1 | Hurt |
| add feature-level mixup to image-level mixup (two-level) (ablation) | MixUp img only · DomainNet Real+Painting sources | average per-class accuracy: 14.3 -> 15.8 | 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.
- MetaReg — MetaReg: Towards Domain Generalization using Meta-Regularization (NeurIPS 2018) · better on 1 of 1 shared result
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
- Distinguishing Unseen From Seen for Generalized Zero-Shot Learning (CVPR 2022 · similarity 0.91)
- Domain-Aware Visual Bias Eliminating for Generalized Zero-Shot Learning (CVPR 2020 · similarity 0.90)
- Rethinking Zero-Shot Learning: A Conditional Visual Classification Perspective (ICCV 2019 · similarity 0.89)
- ZeroDiff: Solidified Visual-semantic Correlation in Zero-Shot Learning (ICLR 2025 · similarity 0.87)
- ZeroMamba: Exploring Visual State Space Model for Zero-Shot Learning (AAAI 2025 · similarity 0.87)
- Improving Generalized Zero-Shot Learning by Exploring the Diverse Semantics from External Class Names (CVPR 2024 · similarity 0.86)
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.
