Research map / Vision-language and multimodal / Cross-modal learning
MOON2.0: Dynamic Modality-balanced Multimodal Representation Learning for E-commerce Product Understanding
Zhanheng Nie, Chenghan Fu, Daoze Zhang, Junxian Wu, Wanxian Guan, Pengjie Wang, Jian Xu, Bo Zheng
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 Modality-driven MoE (vs plain MLP) (ablation) | MOON2.0 · MBE2.0 q_i->c_mm | R@10: 74.59 -> 91.08 | Improved |
| add Modality-driven MoE (vs plain MLP) (ablation) | MOON2.0 · MBE2.0 product classification | accuracy: 62.55 -> 68.08 | Improved |
| add Modality-driven MoE (vs plain MLP) (ablation) | MOON2.0 · MBE2.0 attribute prediction | accuracy: 75.62 -> 84.29 | Improved |
| add Dual-level Alignment (ablation) | MOON2.0 · MBE2.0 q_t->c_i | R@10: 31.45 -> 73.12 | Improved |
| add Dual-level Alignment (ablation) | MOON2.0 · MBE2.0 q_i->c_t | R@10: 23.35 -> 64.91 | Improved |
| add Dual-level Alignment (ablation) | MOON2.0 · MBE2.0 q_t->c_mm | R@10: 37.99 -> 63.09 | Improved |
| add MLLM-based Image-text Co-augmentation (ablation) | MOON2.0 · MBE2.0 q_i->c_mm | R@10: 78.17 -> 91.08 | Improved |
| add MLLM-based Image-text Co-augmentation (ablation) | MOON2.0 · MBE2.0 attribute prediction | accuracy: 77.77 -> 84.29 | Improved |
| add Dynamic Sample Filtering (ablation) | MOON2.0 · MBE2.0 q_mm->c_mm | R@10: 80.00 -> 94.21 | Improved |
| add Dynamic Sample Filtering (ablation) | MOON2.0 · MBE2.0 attribute prediction | accuracy: 84.04 -> 84.29 | No clear effect |
| add Dynamic Sample Filtering (ablation) | MOON2.0 · MBE2.0 product classification | accuracy: 67.99 -> 68.08 | No clear effect |
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.
- MM-Embed — MM-EMBED: UNIVERSAL MULTIMODAL RETRIEVAL WITH MULTIMODAL LLMS (ICLR 2025) · 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
- Beyond Global Similarity: Multi-Conditional Retrieval for Fine-Grained Cross-Modal Understanding (CVPR 2026 · similarity 0.86)
- SAE-V: Interpreting Multimodal Models for Enhanced Alignment (ICML 2025 · similarity 0.85)
- On the Value of Cross-Modal Misalignment in Multimodal Representation Learning (NeurIPS 2025 · similarity 0.85)
- Incorporating Dense Knowledge Alignment into Unified Multimodal Representation Models (CVPR 2025 · similarity 0.85)
- TIGER-FG: Text-Guided Implicit Fine-Grained Grounding for E-commerce Retrieval (NeurIPS 2026 · similarity 0.85)
- Amplifying Prominent Representations in Multimodal Learning via Variational Dirichlet Process (NeurIPS 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.
