AI Paper Atlas

Research map / Vision-language and multimodal / Cross-modal learning

Multimodal Nested Learning for Decoupled and Coordinated Optimization

Yanglin Feng, Yang Qin, Dezhong Peng, Rui Wang, Xiaomin Song, Peng Hu

ICML 2026 · Vision-language and multimodal › Cross-modal learning · approach: Multimodal fusion with missing modalities

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 changedWhereMeasuredEffect
add nested learning structure (vs flat joint optimization) (ablation)MoNet · CREMA-Dfusion accuracy: 63.26 -> 83.11Improved
add different update frequencies (vs ff = fs = 1) (ablation)MoNet · CREMA-Dfusion accuracy: 81.59 -> 83.11Improved
add outer-layer DMSM (ablation)MoNet · CREMA-Dfusion accuracy: 31.14 -> 83.11Improved
add inner-layer AMCF (ablation)MoNet · CREMA-Dfusion accuracy: 79.40 -> 83.11Improved
add CCMS in AMCF (ablation)MoNet · CREMA-Dfusion accuracy: 81.42 -> 83.11Improved
add EMA in AMCF (ablation)MoNet · CREMA-Dfusion accuracy: 82.17 -> 83.11Improved
add AMCF fusion coordination (ablation)MoNet · CREMA-Dunimodal accuracy: Audio 61.88 -> 59.86Mixed
use MoNet vs best baseline DGLMoNet vs DGL · CREMA-Dfusion accuracy: 79.40 -> 83.11Improved
use MoNet vs ARLMoNet vs ARL · VGGSound50fusion accuracy: 60.19 -> 60.90Improved
use MoNet vs InfoRegMoNet vs InfoReg · MVSAfusion accuracy: 74.05 -> 74.25No clear effect
use MoNet vs MMParetoMoNet vs MMPareto · ModelNetfusion accuracy: 91.60 -> 92.50Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy