Research map / Vision-language and multimodal / Cross-modal learning
Hierarchical Multimodal Metric Learning for Multimodal Classification
Check 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 |
|---|---|---|---|
| Use HM3L instead of the strongest reported alternative | 1-NN classification on NUS-WIDE · NUS-WIDE, 1-NN | accuracy: 0.9352 -> 0.9524 | Helped |
| Use HM3L instead of the strongest reported alternative | 3-NN classification on NUS-WIDE · NUS-WIDE, 3-NN | accuracy: 0.9246 -> 0.9431 | Helped |
| Use HM3L instead of the strongest reported alternative | 5-NN classification on NUS-WIDE · NUS-WIDE, 5-NN | accuracy: 0.9299 -> 0.9418 | Helped |
| Use HM3L instead of the strongest reported alternative | 10-NN classification on NUS-WIDE · NUS-WIDE, 10-NN | accuracy: 0.9139 -> 0.9339 | Helped |
| Use HM3L instead of the strongest reported alternative | 20-NN classification on NUS-WIDE · NUS-WIDE, 20-NN | accuracy: 0.9074 -> 0.9223 | Helped |
| Use HM3L instead of MMSS | RGB-D Object instance recognition · RGB-D Object, RGB-D | accuracy: 94.0 -> 95.0 | Helped |
| Use HM3L instead of MMSS | RGB-D Object category recognition · RGB-D Object, RGB-D | accuracy: 88.5 ± 2.2 -> 89.2 ± 1.6 | Helped |
| Use HM3L instead of PMML | CIN 2D3D category recognition · CIN 2D3D, RGB-D | accuracy: 91.8 -> 92.9 | Helped |
| Use HM3L instead of Wang | SUN RGB-D scene categorization · SUN RGB-D, RGB-D | accuracy: 48.1 -> 52.3 | Helped |
Closest papers by meaning
- Rethinking Multimodal Learning from the Perspective of Mitigating Classification Ability Disproportion (NeurIPS 2025 · similarity 0.86)
- Facilitating Multimodal Classification via Dynamically Learning Modality Gap (NeurIPS 2024 · similarity 0.85)
- Multimodal Ensembling for Zero-Shot Image Classification (AAAI 2024 · similarity 0.84)
- Keep the Balance: A Parameter-Efficient Symmetrical Framework for RGB+X Semantic Segmentation (CVPR 2025 · similarity 0.83)
- GeoMM: On Geodesic Perspective for Multi-modal Learning (CVPR 2025 · similarity 0.83)
- With Limited Data for Multimodal Alignment, Let the STRUCTURE Guide You (NeurIPS 2025 · similarity 0.83)
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.