AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

Lumina-OmniLV: A Unified Multimodal Framework for General Low-Level Vision

Yuandong Pu, Le Zhuo, Kaiwen Zhu, Liangbin Xie, Wenlong Zhang, Xiangyu Chen, Peng Gao, Yu Qiao, Chao Dong, Yihao Liu

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Instruction-tuned multimodal assistants

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
use OmniLV instead of GenLV on deblurringGenLV · OLV-T Deblur (6 types)PSNR: 22.15 -> 22.57Improved
use OmniLV instead of PixWizard on depth estimationPixWizard · Depth estimationRMSE: 0.941 -> 0.525Improved
use OmniLV instead of PixWizard on HED edge detectionPixWizard · HED edge detectionMAE: 25.95 -> 20.75Improved
use OmniLV instead of specialized Depth Anything on depthDepth Anything · Depth estimationRMSE: 0.291 -> 0.525Hurt
use OmniLV instead of MAXIM on low-light enhancementMAXIM · LOLv2-Real Low lightPSNR: 34.04 -> 18.60Hurt
use OmniLV instead of GenLV on stylization (LLF)GenLV · LLF stylizationFID: 29.53 -> 24.16Improved
use OmniLV instead of GenLV on stylization (LLF)GenLV · LLF stylizationPSNR: 25.66 -> 21.72Hurt
use OmniLV instead of PixWizard in human preferencePixWizard · Human preference benchmarksElo rating: 1,310 -> 1,475Improved
use OmniLV instead of PixWizard on dense task success rate (Image2Depth)PixWizard · Dense prediction success ratesuccess rate: 93% -> 100%Improved
separate text and visual encoding instead of unified MLLM encoding (ablation)OmniLV · Dense prediction taskstask alignmentImproved
fine-tune base model with adapter (co-training) instead of adapter only (ablation)OmniLV · Condition integration settingsperformanceImproved
use projection-addition instead of input concatenation for visual prompts (ablation)OmniLV · In-context visual promptsperformanceImproved

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