Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Automatic Discovery, Association Estimation and Learning of Semantic Attributes for a Thousand Categories
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 the linguistic prior to regularize class-attribute association weights (ablation) | Joint Model on ImageNet · ImageNet attribute prediction | Accuracy: 55.2 -> 60.3 | Helped |
| add the linguistic prior to regularize class-attribute association weights (ablation) | Joint Model on ImageNet · ImageNet attribute prediction | AP: 22.4 -> 28.9 | Helped |
| add the linguistic prior to regularize class-attribute association weights (ablation) | Joint Model on ImageNet · ImageNet DAP classification | Top1: 30.4 -> 45.2 | Helped |
| add the linguistic prior to regularize class-attribute association weights (ablation) | Joint Model on ImageNet · ImageNet DAP classification | AP: 19.0 -> 39.5 | Helped |
| optimize class-attribute associations using visual data (ablation) | Deep Attribute Model on ImageNet · ImageNet attribute prediction | Accuracy: 74.8 -> 76.9 | Helped |
| optimize class-attribute associations using visual data (ablation) | Deep Attribute Model on ImageNet · ImageNet attribute prediction | AP: 64.1 -> 68.2 | Helped |
| optimize class-attribute associations using visual data (ablation) | Deep Attribute Model on ImageNet · ImageNet DAP classification | Top1: 51.4 -> 55.9 | Helped |
| optimize class-attribute associations using visual data (ablation) | Deep Attribute Model on ImageNet · ImageNet DAP classification | AP: 48.3 -> 54.2 | Helped |
| estimate classifier-score mean and standard deviation using test data as a batch (ablation) | Ours on ImageNet zero-shot split A · ImageNet zero-shot split A | Top5 accuracy: 46.1 -> 48.0; 15.9 -> 20.2 | Helped |
| estimate classifier-score mean and standard deviation using test data as a batch (ablation) | Ours on ImageNet zero-shot split B · ImageNet zero-shot split B | Top5 accuracy: 46.3 -> 49.0; 15.2 -> 20.0 | Helped |
| optimize associations (ablation) | Ours on ImageNet zero-shot split C · ImageNet zero-shot split C | Top5 accuracy: 45.8 -> 48.1; 14.8 -> 16.9 | Helped |
| use continuous rather than binary class-attribute associations (ablation) | Ours on AwA and aPY · AwA zero-shot | Zero-shot performance: 77.3 -> 79.7 | Helped |
Closest papers by meaning
- Compositional Caching for Training-free Open-vocabulary Attribute Detection (CVPR 2025 · similarity 0.88)
- Tree of Attributes Prompt Learning for Vision-Language Models (ICLR 2025 · similarity 0.88)
- Attribute-formed Class-specific Concept Space: Endowing Language Bottleneck Model with Better Interpretability and Scalability (CVPR 2025 · similarity 0.88)
- Black Sheep in the Herd: Playing with Spuriously Correlated Attributes for Vision-Language Recognition (ICLR 2025 · similarity 0.87)
- Class Semantic Attribute Perception Guided Zero-Shot Learning (AAAI 2025 · similarity 0.87)
- GUIDED: Granular Understanding via Identification, Detection, and Discrimination for Fine-Grained Open-Vocabulary Object Detection (NeurIPS 2025 · 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.