Research map / Vision-language and multimodal / Cross-modal learning
Achieving Cross Modal Generalization with Multimodal Unified Representation
Yan Xia, Hai Huang, Jieming Zhu, Zhou Zhao
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 |
|---|---|---|---|
| remove CLUB (ablation) | Uni-Code · VGGsounds-AVEL 24K, AVE V->A | precision: 44.0 -> 34.9 | Hurt |
| remove Cross-CPC (ablation) | Uni-Code · VGGsounds-AVEL 24K, AVE V->A | precision: 44.0 -> 4.6 | Hurt |
| remove MM-EMA (ablation) | Uni-Code · VGGsounds-AVEL 24K, AVE V->A | precision: 44.0 -> 34.1 | Hurt |
| remove reset code (ablation) | Uni-Code · VGGsounds-AVEL 24K, AVE V->A | precision: 44.0 -> 37.8 | Hurt |
| remove Lcmcm loss (ablation) | Uni-Code · VGGsounds-AVEL 24K, AVE V->A | precision: 44.0 -> 39.7 | Hurt |
| remove MM-EMA and reset code (ablation) | Uni-Code · VGGsounds-AVEL 24K, AVE V->A | precision: 44.0 -> 28.2 | Hurt |
| remove Cross-CPC in audio-visual-text pre-training (ablation) | Uni-Code · VGGsounds-AVEL 40K, AVE V->A | precision: 54.1 -> 43.8 | Hurt |
| remove Lcmcm in audio-visual-text pre-training (ablation) | Uni-Code · VGGsounds-AVEL 40K, AVE V->A | precision: 54.1 -> 52.4 | Hurt |
| add DCID to CMCM | CMCM · VGGsounds-AVEL 24K, AVE V->A | precision: 28.9 -> 34.1 | Improved |
| add DCID to CODIS | CODIS · VGGsounds-AVEL 24K, AVE V->A | precision: 18.5 -> 33.4 | Improved |
| increase pre-training data from 24K to 40K | Uni-Code · AVE V->A | precision: 44.0 -> 47.7 | Improved |
| increase pre-training data from 40K to 81K | Uni-Code · AVE V->A | precision: 47.7 -> 41.2 | Hurt |
Closest papers by meaning
- Connect, Collapse, Corrupt: Learning Cross-Modal Tasks with Uni-Modal Data (ICLR 2024 · similarity 0.89)
- Cross-modal Representation Flattening for Multi-modal Domain Generalization (NeurIPS 2024 · similarity 0.88)
- AlignMamba: Enhancing Multimodal Mamba with Local and Global Cross-modal Alignment (CVPR 2025 · similarity 0.88)
- Strong Helps Weak: Directional Cross-Modal Alignment Transfer in Multi-modal LLMs (NeurIPS 2026 · similarity 0.87)
- When to Align, When to Predict: A Phase Diagram for Multimodal Learning (NeurIPS 2026 · similarity 0.87)
- Asymmetric Hierarchical Anchoring for Robust Audio–Visual Cross-Modal Generalization (NeurIPS 2026 · similarity 0.87)
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.
