Research map / Graph learning / GNN architectures
Knowledge Graph Transfer Network for Few-Shot Recognition
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 |
|---|---|---|---|
| transfer classifier information using semantic category correlations | KGTN with ResNet-50 · ImageNet-FS novel 1-shot | Top-5 accuracy: 54.3 -> 62.1 | Helped |
| transfer classifier information using semantic category correlations | KGTN with ResNet-50 · ImageNet-FS all 1-shot | Top-5 accuracy: 60.7 -> 68.3 | Helped |
| add semantic similarity correlations rather than uniform correlations (ablation) | KGTN · ImageNet-FS novel 1-shot | Top-5 accuracy: 53.4 -> 62.1 | Helped |
| add semantic similarity correlations rather than random correlations (ablation) | KGTN · ImageNet-FS novel 1-shot | Top-5 accuracy: 54.4 -> 62.1 | Helped |
| add semantic similarity correlations rather than category hierarchy correlations (ablation) | KGTN · ImageNet-FS novel 1-shot | Top-5 accuracy: 60.1 -> 62.1 | Helped |
| use category hierarchy correlations rather than uniform correlations (ablation) | KGTN · ImageNet-FS novel 1-shot | Top-5 accuracy: 53.4 -> 60.1 | Helped |
| use category hierarchy correlations rather than random correlations (ablation) | KGTN · ImageNet-FS novel 1-shot | Top-5 accuracy: 54.4 -> 60.1 | Helped |
| use cosine similarity instead of inner product (ablation) | KGTN · ImageNet-FS novel 1-shot | Top-5 accuracy: 62.1 -> 61.4 | Hurt |
| use Pearson correlation instead of inner product (ablation) | KGTN · ImageNet-FS novel 1-shot | Top-5 accuracy: 62.1 -> 61.5 | Hurt |
| use cosine similarity instead of inner product (ablation) | KGTN · ImageNet-6K novel 1-shot | Top-5 accuracy: 29.5 -> 30.5 | Helped |
| initialize base classifier nodes randomly instead of with first-stage pretrained parameters (ablation) | KGTN · ImageNet-FS novel 1-shot | Top-5 accuracy: up to 5.0% | Helped |
Closest papers by meaning
- Graph Few-Shot Learning via Knowledge Transfer (AAAI 2020 · similarity 0.92)
- Few-Shot Learning with Graph Neural Networks (ICLR 2018 · similarity 0.90)
- Zero-Shot Recognition via Semantic Embeddings and Knowledge Graphs (CVPR 2018 · similarity 0.90)
- Unlocking the Potential of Black-box Pre-trained GNNs for Graph Few-shot Learning (AAAI 2025 · similarity 0.88)
- Simple Semantic-Aided Few-Shot Learning (CVPR 2024 · similarity 0.86)
- Leveraging Large Language Models for Node Generation in Few-Shot Learning on Text-Attributed Graphs (AAAI 2025 · 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.