Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Joint Intermodal and Intramodal Label Transfers for Extremely Rare or Unseen Classes
Guo-Jun Qi, Wei Liu, Charų C. Aggarwal, Thomas S. Huang
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 |
|---|---|---|---|
| share the label transfer matrix across seen and unseen classes (ablation) | I2LT · CUB200 zero-shot | average AUC: 0.64 -> 0.67 | Improved |
| share the label transfer matrix across seen and unseen classes (ablation) | I2LT · Flowers102 zero-shot | average AUC: 0.70 -> 0.73 | Improved |
| replace BOVW features with CNN features | I2LT · NUS-WIDE I2I search | Precision@20: 56.34 -> 68.72 | Improved |
| replace BOVW features with CNN features | I2LT · NUS-WIDE T2I search | Precision@20: 56.85 -> 69.20 | Improved |
| add joint intermodal and intramodal transfer (I2LT) over TTI | TTI · NUS-WIDE (81 concepts) | MAP: 0.3418 -> 0.4042 | Improved |
| increase the number of training images from two to ten | I2LT · Flickr-Wiki birds | error rate: 0.2523 -> 0.2314 | Improved |
| increase number of training images to a large level | label transfer vs baselines · Flickr-Wiki | error rate | Hurt |
| increase number of text-image co-occurrence pairs | I2LT · Flickr-Wiki | error rate | Improved |
| use I2LT over SVM with two training images | SVM · Flickr-Wiki birds (2 images) | error rate: 0.3293 -> 0.2523 | Improved |
| add the trace norm regularizer for concise transfer model | I2LT · zero-shot CUB200 and Flowers102 | average AUC | Improved |
Closest papers by meaning
- Open-Vocabulary Multi-Label Classification via Multi-Modal Knowledge Transfer (AAAI 2023 · similarity 0.88)
- Heterogeneous Semantic Transfer for Multi-label Recognition with Partial Labels (IJCV 2024 · similarity 0.87)
- Language-Driven Cross-Modal Classifier for Zero-Shot Multi-Label Image Recognition (ICML 2024 · similarity 0.86)
- Semi-Supervised Multimodal Classification Through Learning from Modal and Strategic Complementarities (AAAI 2025 · similarity 0.85)
- Multimodal Ensembling for Zero-Shot Image Classification (AAAI 2024 · similarity 0.85)
- Multi-Label Few-Shot Image Classification via Pairwise Feature Augmentation and Flexible Prompt Learning (AAAI 2025 · 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.
