Research map / 3D vision / 3D generation
Memory Fusion Network for Multi-view Sequential Learning
Read the paperCheck 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 MFN's separate per-view LSTMs, delta-memory attention, and gated fusion memory rather than the strongest prior model | Multi-view sequence learning · CMU-MOSI | BA: 74.6 -> 77.4 | Helped |
| Use MFN rather than the strongest prior model | Multi-view sequence learning · ICT-MMMO | BA: 81.3 -> 87.5 | Helped |
| Use MFN rather than the strongest prior model | Multi-view sequence learning · YouTube | MA(3): 50.2 -> 61.0 | Helped |
| Use MFN rather than the strongest prior model | Multi-view sequence learning · MOUD | BA: 74.0 -> 81.1 | Helped |
| Use MFN rather than the strongest prior model | Multi-view sequence learning · IEMOCAP discrete emotions | MA(9): 36.0 -> 36.5 | Helped |
| Use MFN rather than the strongest prior model | Multi-view sequence learning · POM confident | MA(7): 26.6 -> 34.5 | Helped |
| Add attention over memories at t-1 as well as t (ablation) | MFN · CMU-MOSI | BA: 75.5 -> 77.4 | Helped |
| Add attention over memories at t-1 as well as t (ablation) | MFN · ICT-MMMO | BA: 76.3 -> 87.5 | Helped |
| Add the DMAN and Multi-view Gated Memory (ablation) | MFN · YouTube | MA(3): 47.5 -> 61.0 | Helped |
| Add the DMAN and Multi-view Gated Memory (ablation) | MFN · IEMOCAP discrete emotions | MA(9): 31.2 -> 36.5 | Helped |
| Use multiple views rather than only the language view (ablation) | MFN · CMU-MOSI | BA: 73.2 -> 77.4 | Helped |
| Expand DMAN's input region from [t-1,t] to [t-q,t] (ablation) | MFN · Experiments across datasets | performance | No clear effect |
Closest papers by meaning
- Fundamental limits of learning in sequence multi-index models and deep attention networks: high-dimensional asymptotics and sharp thresholds (ICML 2025 · similarity 0.85)
- SeqMvRL: A Sequential Fusion Framework for Multi-view Representation Learning (CVPR 2025 · similarity 0.84)
- Multi-View Incremental Learning with Structured Hebbian Plasticity for Enhanced Fusion Efficiency (AAAI 2025 · similarity 0.83)
- Towards flexible perception with visual memory (ICML 2025 · similarity 0.83)
- Swiss Army Knife: Synergizing Biases in Knowledge from Vision Foundation Models for Multi-Task Learning (ICLR 2025 · similarity 0.83)
- Understanding the Differences in Foundation Models: Attention, State Space Models, and Recurrent Neural Networks (NeurIPS 2024 · 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.