AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

RegRet: Enhancing Region-Level Retrieval in Large Multimodal Models

Xun Liang, Honghui Yang, Weihang Pan, Boyuan Pan, Yao Hu, Binbin Lin, Deng Cai, Ruisi Zhao, Wenxiao Wang

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Instruction-tuned multimodal assistants · 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 layer-wise coordination between CE and RAE features (ablation)RegRet-8B RAE · REGMBaverage recall: 82.6 -> 86.2Improved
use separate projectors for RAE and context encoder (ablation)RegRet-8B RAE · REGMBaverage recall: 66.0 -> 82.6Improved
use separate projectors for RAE and context encoder (ablation)RegRet-8B RAE · M-BEIRaverage recall: 27.2 -> 57.1Improved
replace auxiliary-image prompting with RAE (ablation)RegRet-8B · REGMBaverage recall: 83.4 -> 86.2Improved
add regional contrastive training data (REGMB) to 3B model (ablation)RegRet-3B · REGMBaverage recall: 69.4 -> 79.9Improved
add image-level (M-BEIR) data to text-only training (ablation)RegRet-3B · REGMBaverage recall: 68.4 -> 69.4Mixed
train LamRA on REGMB (ablation)LamRA · REGMBaverage recall: 67.7 -> 80.9Improved
add regional contrastive training to zero-shot RegRet-8BRegRet-8B · Public region benchmarksaverage: 48.3 -> 56.1Improved
use RAE instead of prompting strategies (crop, visual mark, auxiliary image) on LamRA (ablation)zero-shot REGMB · REGMBaverage recall: 63.1 -> 78.8Improved
mixed visual prompting in Stage III (RAE-only for tasks 1-3, CE+RAE for task 4) (ablation)RegRet-3B · REGMBaverage recall: 78.4 -> 79.9Improved
train RegRet on region data, global retrievalRegRet-8B vs LamRA · M-BEIRaverage recall: 57.1 -> 57.3No clear effect
RegRet-8B vs RzenEmbed on context-level xgoods taskRegRet-8B · RegMB Task 4 xgoodsR@5: 98.2 -> 93.6Hurt

Earlier methods this paper compares with

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