AI Paper Atlas

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

Robust Multimodal Learning via Representation Decoupling

Shicai Wei, Yang Luo, Yuji Wang, Chunbo Luo

ECCV 2024 · Vision-language and multimodal › Cross-modal learning · approach: Multimodal fusion with missing modalities · 26 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 decoupled multimodal representation (sample embeddings from learned distribution) (ablation)SF-MD baseline · CASIA-SURFaverage ACER: 7.52 -> 4.63Improved
add hard combination regularizer on top of DMR (ablation)DMRNet · CASIA-SURFaverage ACER: 4.63 -> 3.58Improved
estimate distribution on feature map via convolution instead of feature-vector PE or PCME (ablation)DMR · CASIA-SURFaverage ACER: 5.85 -> 4.66Improved
set distributional regularization weight alpha to 0 (ablation)DMR · CASIA-SURFaverage ACER: 4.66 -> 8.18Hurt
set alpha too large (1e-1) (ablation)DMR · CASIA-SURFaverage ACER: 4.66 -> 49.76Hurt
set hard combination weight beta to 0.7 instead of 0.1 or 0.9 (ablation)DMRNet · CASIA-SURFaverage ACER: 4.31 -> 3.58Improved
use DMRNet instead of ESANet-MD for multimodal segmentationESANet · NYUv2average mIoU: 42.77 -> 45.08Improved
use DMRNet instead of ShaSpec on audio-only CREMA-DShaSpec · CREMA-D (audio only)accuracy: 59.86 -> 58.87Hurt
use DMRNet instead of MMANet on CASIA-SURFMMANet · CASIA-SURFaverage ACER: 3.94 -> 3.58Improved
use DMRNet instead of ShaSpec and MMANet on Kinetics-SoundsShaSpec · Kinetics-Soundsaverage accuracy: 51.62 -> 56.01Improved

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