AI Paper Atlas

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

Cross-Modal Knowledge Distillation without Paired Data: Theoretical Foundation and Algorithm

T. Khiem Tran, Anh Duc Chu, Quang Hung Pham, Phi Le Nguyen, Trong Nghia Hoang

ICML 2026 · Vision-language and multimodal › Cross-modal learning · approach: Unimodal-to-multimodal representation transfer · 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
add bi-level optimization to joint feature and label alignment (ablation)UCMKD · RAVDESS A→Vaccuracy (%): 66.47 -> 73.83Improved
add bi-level optimization to joint feature and label alignment (ablation)UCMKD · AVE A→Vaccuracy (%): 28.11 -> 34.16Improved
add label alignment to feature alignment (ablation)UCMKD · RAVDESS A→Vaccuracy (%): 65.37 -> 73.83Improved
add feature alignment to label alignment (ablation)UCMKD · AVE V→Aaccuracy (%): 48.92 -> 52.24Improved
use l2 instead of l1 transport cost (ablation)UCMKD · RAVDESS A→Vaccuracy (%): 74.23 vs 73.83No clear effect
use l2 instead of angular transport cost (ablation)UCMKD · CREMA-D V→Aaccuracy (%): 62.41 -> 66.59Improved
use UCMKD unpaired instead of vanilla KD pairedUCMKD · AVE A→Vaccuracy (%): 29.85 -> 34.16Improved
use UCMKD unpaired instead of vanilla KD pairedUCMKD · CREMA-D A→Vaccuracy (%): 72.18 -> 71.64Hurt
use UCMKD instead of Cross EntropyUCMKD · RAVDESS A→V (50% data)accuracy (%): 59.87 -> 72.76Improved
use UCMKD instead of Cross EntropyUCMKD · RAVDESS A→V (30% data)accuracy (%): 54.51 -> 69.03Improved
use UCMKD instead of Cross Entropy with ViT backboneUCMKD · RAVDESS A→V ViT-Saccuracy (%): 65.63 -> 80.32Improved
use UCMKD instead of C2KD in paired settingUCMKD · RAVDESS V→A pairedaccuracy (%): 82.78 -> 75.13Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 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