AI Paper Atlas

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

Boosting Multimodal Learning via Disentangled Gradient Learning

Shicai Wei, Chunbo Luo, Yang Luo

ICCV 2025 · Vision-language and multimodal › Cross-modal learning · approach: Multimodal fusion with missing modalities · 23 citations (OpenAlex)

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 multimodal gradient truncation (MT) (ablation)concat multimodal baseline · CREMA-Daccuracy: 65.10 -> 73.31Improved
add unimodal gradient truncation (UT) (ablation)concat multimodal baseline · CREMA-Daccuracy: 65.10 -> 74.22Improved
combine MT and UT (full DGL) (ablation)concat multimodal baseline · CREMA-Daccuracy: 65.10 -> 77.48Improved
raise unimodal loss weight alpha from 1 to 4 (ablation)DGL · CREMA-Daccuracy: 73.35 -> 77.48Improved
raise alpha from 4 to 5 (ablation)DGL · CREMA-Daccuracy: 77.48 -> 77.08Hurt
raise alpha from 1 to 4 (ablation)G-Blending · CREMA-Daccuracy: 67.89 -> 69.21Improved
replace concat baseline with DGLConcat · VGGSoundmultimodal accuracy: 47.90 -> 52.53Improved
replace concat baseline with DGLConcat · Kinetics-Soundsmultimodal accuracy: 64.45 -> 74.78Improved
apply DGL to MMTM fusionMMTM multimodal baseline · CREMA-Dmultimodal accuracy: 60.12 -> 75.00Improved
apply DGL to three-modality MLP fusionMultimodal baseline · MOSImultimodal accuracy: 76.83 -> 79.78Improved
apply DGL to ESANet (from scratch)ESANet · NYUv2mIoU: 38.59 -> 41.67Improved
apply DGL to ESANet (ImageNet pretrained)ESANet · NYUv2mIoU: 48.48 -> 50.10Improved

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