Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Attribute Prototype Network for Any-Shot Learning
Wenjia Xu, Yongqin Xian, Jiuniu Wang, Bernt Schiele, Zeynep Akata
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 attribute regression loss L_Reg with ProtoMod (ablation) | BaseMod · CUB ZSL | top-1 accuracy: 70.0 -> 71.5 | Improved |
| add attribute decorrelation loss L_AD (ablation) | BaseMod + L_Reg · AWA2 ZSL | top-1 accuracy: 66.3 -> 67.7 | Improved |
| add attribute decorrelation loss L_AD (ablation) | BaseMod + L_Reg · CUB part localization | PCP mean: 35.2 -> 49.0 | Improved |
| add compactness regularizer L_CPT (ablation) | BaseMod + L_Reg + L_AD · ZSL CUB, AWA2, SUN | top-1 accuracy: 71.8 -> 72.0 | No clear effect |
| add compactness regularizer L_CPT (ablation) | BaseMod + L_Reg + L_AD · CUB part localization | PCP mean: 49.0 -> 52.8 | Improved |
| add ZoomInMod (ablation) | APN without ZoomInMod · CUB ZSL | top-1 accuracy: 72.0 -> 75.0 | Improved |
| add ZoomInMod (ablation) | APN without ZoomInMod · SUN ZSL | top-1 accuracy: 61.6 -> 61.5 | No clear effect |
| add ZoomInMod (ablation) | APN · compute cost | FLOPs (GMac): 7.9 -> 15.7 | Hurt |
| use APN features instead of finetuned features for ABP | generative ZSL model · AWA2 ZSL | top-1 accuracy: 68.5 -> 73.8 | Improved |
| use APN features instead of finetuned features for CVC | generative ZSL model · AWA2 ZSL | top-1 accuracy: 64.6 -> 71.2 | Improved |
| use APN features for ABP | generative ZSL model · SUN ZSL | top-1 accuracy: 62.6 -> 63.1 | Improved |
| use APN features for CVC | generative GZSL model · AWA2 GZSL unseen / seen | accuracy: gains 5.8% on unseen while sacrificing only 2.1% on seen | Improved |
Closest papers by meaning
- Attribute Prototype Network for Zero-Shot Learning (NeurIPS 2020 · similarity 0.97)
- TransZero: Attribute-Guided Transformer for Zero-Shot Learning (AAAI 2022 · similarity 0.91)
- TransZero++: Cross Attribute-Guided Transformer for Zero-Shot Learning (TPAMI 2022 · similarity 0.90)
- Fine-Grained Generalized Zero-Shot Learning via Dense Attribute-Based Attention (CVPR 2020 · similarity 0.89)
- Disentangling Visual Embeddings for Attributes and Objects (CVPR 2022 · similarity 0.89)
- Class Semantic Attribute Perception Guided Zero-Shot Learning (AAAI 2025 · 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.
