Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Intervention-Guided Image-Free Classifier Expansion for Fine-Grained Recognition
Xiangyu Wang, Changxin Rong, Yanze Gao, Lyuzhou Chen, Xiren Zhou, Huanhuan Chen
Read the paperCheck your idea against 137,217 papers
AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes an intervention-guided approach to image-free zero-shot classifier expansion that targets fine-grained recognition of visually and semantically similar classes.
Problem
Image-free zero-shot classifier expansion must extend recognition to unseen classes by mapping class semantics to visual classifier weights without original images; existing association-based optimization struggles with fine-grained, visually and semantically similar classes.
Approach
Replaces purely association-based optimization with an intervention-guided mechanism to better separate visually and semantically similar classes; full details are unavailable beyond the abstract.
Closest papers by meaning
- Granularity-Aware Adaptive Classifier Expansion via Zero-Shot Learning (ICML 2026 · similarity 0.89)
- Counterfactual-Driven Zero-Shot Classifier Expansion (AAAI 2026 · similarity 0.88)
- Image-Free Classifier Injection for Zero-Shot Classification (ICCV 2023 · similarity 0.87)
- Class Semantic Attribute Perception Guided Zero-Shot Learning (AAAI 2025 · similarity 0.87)
- Hierarchical Disentanglement of Discriminative Latent Features for Zero-Shot Learning (CVPR 2019 · similarity 0.86)
- Fine-Grained Generalized Zero-Shot Learning via Dense Attribute-Based Attention (CVPR 2020 · 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.
