AI Paper Atlas

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

ELVA: Exploring Ranking-Driven Universal Multimodal Retrieval

Yuhan Liu, Pei Fu, Hang Li, Yukun Qi, Chao Jiang, Jingwen Fu, Zhen Liu, Bin Qin, Zhenbo Luo, Jian Luan, Jingmin Xin

ECCV 2026 · Vision-language and multimodal › Cross-modal learning · approach: Probabilistic embeddings for cross-modal retrieval · 7 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 ranking reward (ablation)ELVA-7B · M-BEIR image-to-text subset averageRecall: 57.2 -> 58.7Improved
add margin reward (ablation)ELVA-7B · M-BEIR image-to-text subset averageRecall: 58.1 -> 58.7Improved
add negative-ranking term to the ranking reward (ablation)ELVA-7B · M-BEIR image-to-text subset averageRecall: 58.1 -> 58.7Improved
add balanced negative sampling strategy (ablation)ELVA-7B · M-BEIR image-to-text subset averageRecall: 58.2 -> 58.7Improved
add randomly sampled negatives (ablation)ELVA-7B · M-BEIR image-to-text subset averageRecall: 58.4 -> 58.7Improved
weight ranking reward higher (alpha 0.4, epsilon 0.6 vs 0.6/0.4) (ablation)ELVA-7B · M-BEIR image-to-text subset averageRecall: 58.2 -> 58.7Improved
weight ranking reward higher (alpha 0.4, epsilon 0.6 vs 0.5/0.5) (ablation)ELVA-7B · M-BEIR image-to-text subset averageRecall: 58.4 -> 58.7Improved
apply ELVA ranking RL to PUMAPUMA · M-BEIRaverage Recall: 54.4 -> 56.3Improved
apply ELVA ranking RL to MM-EmbedMM-Embed · M-BEIRaverage Recall: 52.7 -> 54.9Improved
apply ELVA ranking RL to PUMAPUMA · MRBenchaverage Recall: 35.1 -> 37.0Improved
apply ELVA ranking RL to MM-EmbedMM-Embed · MRBenchaverage Recall: 34.6 -> 36.2Improved
ELVA RL stage widens embedding distance when a grain is maskedELVA · MRBenchcosine distance: 0.07 -> 0.15Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 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