Research map / Vision-language and multimodal / CLIP and zero-shot
DUET: Cross-Modal Semantic Grounding for Contrastive Zero-Shot Learning
Read the paperCheck 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 attribute-level contrastive learning to CSG (ablation) | DUET · AWA2 | H accuracy: +3.5% | Helped |
| Add a trainable language encoder instead of freezing it (ablation) | DUET CSG · AWA2 | H accuracy: 66.5 -> 69.2 | Helped |
| Use semi-serialized attributes with prompts rather than only prompts (ablation) | DUET CSG · AWA2 | H accuracy: 61.7 -> 69.2 | Helped |
| Add prompts to the attribute sequence (ablation) | DUET CSG · AWA2 | H accuracy: 64.9 -> 69.2 | Helped |
| Use linear weighted target-attribute sampling rather than random sampling (ablation) | DUET CSG · AWA2 | H accuracy: 66.9 -> 69.2 | Helped |
| Prune attributes during training (ablation) | DUET CSG · AWA2 | H accuracy: 67.4 -> 69.2 | Helped |
| Add class-level contrastive learning (ablation) | DUET CSG · AWA2 | H accuracy: 68.4 -> 69.2 | Helped |
| Add cross-modal semantic grounding and attribute-level contrastive learning to the vision encoder (ablation) | DUET · AWA2 | H accuracy: 64.1 -> 72.7 | Helped |
| Use DUET instead of TransZero | zero-shot image classification · CUB GZSL | H accuracy: 68.8 -> 67.5 | Hurt |
| Use DUET instead of HSVA | zero-shot image classification · SUN GZSL | H accuracy: 43.3 -> 45.8 | Helped |
| Use DUET instead of the strongest reported baseline | zero-shot image classification · AWA2 GZSL | H accuracy: +1.2% | Helped |
Closest papers by meaning
- MSDN: Mutually Semantic Distillation Network for Zero-Shot Learning (CVPR 2022 · similarity 0.92)
- Class Semantic Attribute Perception Guided Zero-Shot Learning (AAAI 2025 · similarity 0.91)
- ZeroDiff: Solidified Visual-semantic Correlation in Zero-Shot Learning (ICLR 2025 · similarity 0.91)
- Progressive Semantic-Guided Vision Transformer for Zero-Shot Learning (CVPR 2024 · similarity 0.90)
- ZeroMamba: Exploring Visual State Space Model for Zero-Shot Learning (AAAI 2025 · similarity 0.89)
- Interpreting and Analysing CLIP's Zero-Shot Image Classification via Mutual Knowledge (NeurIPS 2024 · similarity 0.89)
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.