Research map / Vision-language and multimodal / Cross-modal learning
Object-Oriented Anchoring and Modal Alignment in Multimodal Learning
Shibin Mei, Bingbing Ni, Hang Wang, Chenglong Zhao, fengfa hu, Zhiming Pi, BiLian Ke
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 region object anchors (ablation) | Base Model · COCO zero-shot text retrieval | R@1: 68.92 -> 70.80 | Improved |
| add region object anchors (ablation) | Base Model · SNLI-VE val | accuracy: 77.03 -> 79.29 | Improved |
| add optimal transport alignment (ablation) | Base + region object · NLVR2 dev | accuracy: 77.98 -> 79.08 | Improved |
| add multiview text mining (ablation) | Base + region + OT · NLVR2 dev | accuracy: 79.08 -> 80.15 | Improved |
| add single-modal DINO supervision (ablation) | Base + region + OT + multiview · COCO zero-shot text retrieval | R@1: 71.59 -> 71.74 | Improved |
| add all four components to base model (ablation) | ALBEF-style Base Model · VQA test-dev | accuracy: 72.78 -> 75.16 | Improved |
| scale pre-training from 4M to 13M images | Ours · COCO text retrieval | R@1: 76.3 -> 80.3 | Improved |
| use 13M data versus BLIP 14M | Ours(13M) · COCO text retrieval | R@1: 80.3 vs 80.6 | Mixed |
Closest papers by meaning
- With Limited Data for Multimodal Alignment, Let the STRUCTURE Guide You (NeurIPS 2025 · similarity 0.90)
- A TRIANGLE Enables Multimodal Alignment Beyond Cosine Similarity (NeurIPS 2025 · similarity 0.89)
- DeepAlign: Mitigating Modality Conflict through Modality-Specific Alignment (CVPR 2026 · similarity 0.89)
- Implicit Multimodal Alignment: On the Generalization of Frozen LLMs to Multimodal Inputs (NeurIPS 2024 · similarity 0.89)
- Strong Helps Weak: Directional Cross-Modal Alignment Transfer in Multi-modal LLMs (NeurIPS 2026 · similarity 0.89)
- Gramian Multimodal Representation Learning and Alignment (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.
