Research map / Video / Video-language understanding
VBF++: Variational Bayesian Fusion with Context-Aware Priors and Recommendation-Guided Adversarial Refinement for Multimodal Video Recommendation
Ziyi Cao, Rui Liu, Yong Chen
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 the variational encoder (probabilistic fusion) (ablation) | VBF++ · TikTok | P@10: 0.1523 -> 0.1635 | Improved |
| add context-aware mixture prior (ablation) | VBF++ · MovieLens | P@10: 0.0593 -> 0.0627 | Improved |
| add context-aware mixture prior (ablation) | VBF++ · TikTok | P@10: 0.1567 -> 0.1635 | Improved |
| add Recommendation-guided Adversarial Refinement (ablation) | VBF++ · MovieLens | P@10: 0.0612 -> 0.0627 | Improved |
| add RAR (ablation) | VBF++ · TikTok | P@10: 0.1598 -> 0.1635 | Improved |
| add meta-learning (ablation) | VBF++ · MovieLens | P@10: 0.0609 -> 0.0627 | Improved |
| add meta-learning (ablation) | VBF++ · Kwai | P@10: 0.1398 -> 0.1423 | Improved |
| replace GCN fixed fusion with full VBF++ (ablation) | VBF++ · Kwai | P@10: 0.1245 -> 0.1423 | Improved |
| sample multiple fusion strategies at inference for high-uncertainty videos | VBF++ · top 20% uncertain videos | hit rate: +3.2% | Improved |
| use VBF++ instead of MVideoRec in cross-domain 50-shot adaptation | VBF++ · TikTok cross-domain (Music/Edu) | P@10: 0.0147 -> 0.0189 (Music); 0.0168 -> 0.0203 (Edu) | Improved |
Closest papers by meaning
- Curriculum Conditioned Diffusion for Multimodal Recommendation (AAAI 2025 · similarity 0.85)
- MENTOR: Multi-level Self-supervised Learning for Multimodal Recommendation (AAAI 2025 · similarity 0.84)
- Beyond Graph Convolution: Multimodal Recommendation with Topology-aware MLPs (AAAI 2025 · similarity 0.84)
- Calibrating Video Watch-time Predictions with Credible Prototype Alignment (ICML 2025 · similarity 0.84)
- In-context Prompt-augmented Micro-video Popularity Prediction (AAAI 2025 · similarity 0.83)
- Structured Spectral Reasoning for Frequency-Adaptive Multimodal Recommendation (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.
