AI Paper Atlas

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

Calibrated Multimodal Representation Learning with Missing Modalities

Xiaohao Liu, Xiaobo Xia, Jiaheng Wei, Shuo Yang, Xiu Su, See-Kiong Ng, Tat-Seng Chua

ICML 2026 · Vision-language and multimodal › Cross-modal learning · approach: Multimodal fusion with missing modalities · 0 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
impute missing modalities at representation level with generative model (CalMRL) on top of PMRLPMRL with VAST backbone · 6 retrieval benchmarks, avg Recall@1avg Recall@1: 53.8 -> 54.2Improved
add CalMRL calibration for classificationPMRL with ImageBind backbone · VGGSound, UCF101, AudioSet, ESC50 avg accuracyaccuracy: 44.04 -> 45.19Improved
continual training on missing-modality datasetsVAST · 6 retrieval benchmarks, avg Recall@1avg Recall@1: 44.8 -> 53.3Improved
train on sole audio-text dataset without calibration (ablation)VAST · DiDeMo T→V and ActivityNet T→VRecall@1: -3.6 and -2.9 vs VASTHurt
add CalMRL calibration on sole audio-text training (ablation)PMRL with sole audio-text training · 6 retrieval benchmarks, avg Recall@1avg Recall@1: 51.2 -> 51.7Improved
add CalMRL calibration on sole video-text training (ablation)PMRL with sole video-text training · 6 retrieval benchmarks, avg Recall@1avg Recall@1: 50.0 -> 50.4Improved
replace linear-Gaussian generative imputation with MVAE (ablation)CalMRL · 6 retrieval benchmarks, avg Recall@1avg Recall@1: 54.8 -> 54.0Hurt
replace generative imputation with MoPoE (ablation)CalMRL · 6 retrieval benchmarks, avg Recall@1avg Recall@1: 54.8 -> 54.1Hurt
replace generative imputation with SMIL or Knowledge Bridger (ablation)CalMRL · 6 retrieval benchmarks, avg Recall@1avg Recall@1: 54.8 -> 54.0 and 54.2Hurt
remove instance matching loss (alpha=0) (ablation)CalMRL · 6 retrieval benchmarks, avg Recall@1avg Recall@1: 54.80 -> 53.30Hurt
over-weight instance matching loss (alpha=1) (ablation)CalMRL · 6 retrieval benchmarks, avg Recall@1avg Recall@1: 54.80 -> 53.21Hurt
vary temperature tau from 0.05 to 0.1-1 (ablation)CalMRL · 6 retrieval benchmarks, avg Recall@1avg Recall@1: 54.80 -> 54.37No clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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