AI Paper Atlas

Research map / Vision-language and multimodal / Human and embodied modeling

LOREAL: Mitigating Low-Resolution Challenges in Vision-Language Models with Attribute-driven Prompt Self-Distillation

Xucong Wang, Pengkun Wang, Zhe Zhao, Liheng Yu, Rui Mao, Yang Wang

CVPR 2026 · Vision-language and multimodal › Human and embodied modeling · approach: Vision-language pretraining

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 LOREAL to MMA at 96x96MMA · LR-B2N average, 96x96HM: 40.50 -> 63.14Improved
add LOREAL to MMRL at 96x96MMRL · LR-B2N average, 96x96HM: 38.90 -> 61.71Improved
add LOREAL to CoOp at 96x96CoOp · LR-B2N average, 96x96HM: 35.54 -> 47.48Improved
add LOREAL near standard resolution (192x192)MMA · LR-B2N average, 192x192HM: 78.27 -> 79.70Improved
add LOREAL to MMRLMMRL · LR-CE target average, 96x96accuracy: 40.77 -> 55.08Improved
add LOREAL to MMRLMMRL · LR-DG target average, 96x96accuracy: 15.42 -> 30.64Improved
add LOREAL, extra cost of meta-netsMaPLe · 96x96 efficiencytunable parameters / HM: +104.5K params, +4ms train, HM 34.85 -> 57.25Improved
add LR to standard-resolution student direction only vs both directions (ablation)LOREAL · LR-B2N average, 144x144HM: 68.91 -> 69.78Improved
add LLD and HLD distillation losses (ablation)LOREAL · LR-B2N average, 144x144HM: 69.78 -> 71.73Improved
low-level distillation is more important than high-level distillation (ablation)LOREAL · LR-B2N average, 144x144HMImproved
increase meta-net dimension Ds from 32 to 2048 (ablation)LOREAL · LR-B2N average, 144x144HM: 71.73 -> 67.60Hurt
increase attribute prompt length M (ablation)LOREAL · LR-B2NHM: about -2%Hurt

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