Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Granularity-Aware Adaptive Classifier Expansion via Zero-Shot Learning
Xiangyu Wang, Yanze Gao, Changxin Rong, Lyuzhou Chen, Xiren Zhou, Huanhuan Chen
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 |
|---|---|---|---|
| replace shared semantic-to-weight mapping with granularity-aware adaptive framework (CSD + ACS) (ablation) | zero-shot classifier expansion · CUB (ZSL) | top-1 accuracy T1 (%): 59.2 -> 68.2 (single semantics) | Improved |
| add multi-source semantics generation (MSG) to a shared mapping (ablation) | shared mapping baseline · CUB (ZSL) | top-1 accuracy T1 (%): 59.2 -> 63.1 | Improved |
| add granularity modeling to multi-source semantics (ablation) | MSG · AWA2 (ZSL) | top-1 accuracy T1 (%): 62.6 -> 77.2 | Improved |
| add granularity-aware modeling vs richer semantics: average gains (ablation) | 2x2 attribution · CUB, AWA2, SUN | average gain: semantics 3.8% vs granularity 8.3% | Improved |
| add fine-grained adapters without structural priors (no CSD) (ablation) | framework · ablation (Figure 3) | T1 and H | Improved |
| introduce CSD module as routing priors (ablation) | framework · ablation (Figure 3) | T1 and H | Improved |
| add causal invariance loss and multi-source semantics in ACS (ablation) | framework · ablation (Figure 3) | T1 and H | Improved |
| give baselines the MSG semantics as input | ICIS · CUB (GZSL) | harmonic mean H (%): 57.2±1.0 -> 59.4±0.9 | Improved |
| give baselines the MSG semantics as input | ICIS · SUN (GZSL) | harmonic mean H (%): 35.1±2.9 -> 33.6±2.5 | Hurt |
| give baselines the MSG semantics as input | CEMIL · CUB (GZSL) | harmonic mean H (%): 60.7±0.7 -> 60.9±0.8 | No clear effect |
| use proposed method instead of CEMIL under distribution shift from ImageNet | zero-shot classifier expansion · AWA2 (Wiki2Vec, ZSL) | top-1 accuracy T1 (%): 73.3±1.2 -> 83.1±0.9 | Improved |
| use proposed method in class-incremental expansion | zero-shot classifier expansion · AWA2 (5 steps) | harmonic mean H (%): 59.7±2.5 -> 66.7±2.1 at step 5 | Improved |
Closest papers by meaning
- Intervention-Guided Image-Free Classifier Expansion for Fine-Grained Recognition (NeurIPS 2026 · similarity 0.89)
- Counterfactual-Driven Zero-Shot Classifier Expansion (AAAI 2026 · similarity 0.88)
- Class Distribution Shifts in Zero-Shot Learning: Learning Robust Representations (NeurIPS 2024 · similarity 0.86)
- Bidirectional Logits Tree: Pursuing Granularity Reconcilement in Fine-Grained Classification (AAAI 2025 · similarity 0.85)
- Class Semantic Attribute Perception Guided Zero-Shot Learning (AAAI 2025 · similarity 0.85)
- Improving Generalized Zero-Shot Learning by Exploring the Diverse Semantics from External Class Names (CVPR 2024 · similarity 0.85)
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.
