Research map / Vision-language and multimodal / Cross-modal learning
GraphAdapter: Tuning Vision-Language Models With Dual Knowledge Graph
Xin Li, Dongze Lian, Zhihe Lu, Jiawang Bai, Zhibo Chen, Xinchao Wang
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 |
|---|---|---|---|
| add fused textual and visual knowledge graphs to the textual adapter | CLIP ResNet-50 · 11-dataset few-shot classification, 16-shot (average) | accuracy: 76.22 vs 75.65 | Improved |
| add fused textual and visual knowledge graphs to the textual adapter | CLIP ResNet-50 · FGVCAircraft, 16-shot | accuracy: 36.87 vs 35.86 | Improved |
| add textual knowledge sub-graph (ablation) | GraphAdapter-T · ImageNet, 16-shot | accuracy: 64.30% -> 65.70 | Improved |
| add visual knowledge sub-graph (ablation) | GraphAdapter-T · ImageNet, 16-shot | accuracy: 64.95% -> 65.70% | Improved |
| move GraphAdapter from the textual branch to the visual branch (ablation) | CLIP ResNet-50 · 11-dataset few-shot classification, 16-shot (average) | accuracy: 76.22 -> 72.21 | Hurt |
| apply GraphAdapter to both textual and visual branches rather than the textual branch alone (ablation) | CLIP ResNet-50 · 11-dataset few-shot classification, 16-shot (average) | accuracy: 76.22 -> 76.25 | No clear effect |
| use learnable α and β instead of selected fixed coefficients (ablation) | GraphAdapter-T · ImageNet few-shot classification | accuracy | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- Tip-Adapter — Tip-Adapter: Training-Free Adaption of CLIP for Few-Shot Classification (ECCV 2022) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- p-Laplacian Adaptation for Generative Pre-trained Vision-Language Models (AAAI 2024 · similarity 0.90)
- MMA: Multi-Modal Adapter for Vision-Language Models (CVPR 2024 · similarity 0.90)
- Hawaii: Hierarchical Visual Knowledge Transfer for Efficient Vision-Language Models (NeurIPS 2025 · similarity 0.89)
- VMT-Adapter: Parameter-Efficient Transfer Learning for Multi-Task Dense Scene Understanding (AAAI 2024 · similarity 0.89)
- GraphVLM: Benchmarking Vision Language Models for Multimodal Graph Learning (CVPR 2026 · similarity 0.89)
- Efficient and Context-Aware Label Propagation for Zero-/Few-Shot Training-Free Adaptation of Vision-Language Model (ICLR 2025 · similarity 0.88)
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.
