Research map / Vision-language and multimodal / CLIP and zero-shot
Stacked Semantics-Guided Attention Model for Fine-Grained 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 |
|---|---|---|---|
| Use S²GA with ground-truth part regions instead of the strongest listed baseline | Fine-grained zero-shot classification on CUB · CUB, attribute and Word2Vec | accuracy (%): 62.0 -> 75.3; 32.1 -> 46.9 | Helped |
| Use S²GA-DET instead of the strongest listed baseline | Fine-grained zero-shot classification · NABird, SCS and SCE splits | accuracy (%): 35.6 -> 39.4; 8.6 -> 9.7 | Helped |
| Use S²GA-DET instead of GAA | Fine-grained zero-shot classification · CUB, SCS split | accuracy (%): 43.7 -> 42.9 | Hurt |
| Use S²GA-DET instead of GAA | Fine-grained zero-shot classification · CUB, SCE split | accuracy (%): 10.3 -> 10.9 | Helped |
| Use S²GA instead of GAA | Zero-shot retrieval · CUB, 50% images per class | mAP (%): 48.3 -> 47.1 | Hurt |
| Use S²GA instead of GAA | Zero-shot retrieval · CUB, 100% images per class | mAP (%): 40.3 -> 42.6 | Helped |
| Use S²GA instead of GAA | Zero-shot retrieval · NABird, 50% and 100% images per class | mAP (%): 37.8 -> 42.2; 31.0 -> 36.6 | Helped |
| Add attention rather than use the no-attention baseline (ablation) | S²GA · CUB and NABird, classification | accuracy (%): 62.5 -> 68.9; 38.4 -> 41.8; 38.6 -> 42.9; 31.6 -> 39.4 | Helped |
| Add a second attention layer instead of using one (ablation) | S²GA · CUB and NABird, classification | accuracy (%): 67.1 -> 68.9; 40.3 -> 41.8; 41.8 -> 42.9; 36.2 -> 39.4 | Helped |
| Add a third attention layer instead of using two (ablation) | S²GA · CUB, attribute, Word2Vec, and TF-IDF; NABird, TF-IDF | accuracy (%): 68.9 -> 68.5; 41.8 -> 41.6; 42.9 -> 42.7; 39.4 -> 39.6 | Mixed |
Closest papers by meaning
- Epsilon: Exploring Comprehensive Visual-Semantic Projection for Multi-Label Zero-Shot Learning (AAAI 2025 · similarity 0.89)
- Class Semantic Attribute Perception Guided Zero-Shot Learning (AAAI 2025 · similarity 0.88)
- ZeroMamba: Exploring Visual State Space Model for Zero-Shot Learning (AAAI 2025 · similarity 0.88)
- Progressive Semantic-Guided Vision Transformer for Zero-Shot Learning (CVPR 2024 · similarity 0.88)
- Data-Free Generalized Zero-Shot Learning (AAAI 2024 · similarity 0.88)
- Visual and Semantic Prompt Collaboration for Generalized Zero-Shot Learning (CVPR 2025 · similarity 0.87)
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.