AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

LIRA: Inferring Segmentation in Large Multi-modal Models with Local Interleaved Region Assistance

Zhang Li, Biao Yang, Qiang Liu, Shuo Zhang, Zhiyin Ma, Liang Yin, Linger Deng, Yabo Sun, Yuliang Liu, Xiang Bai

ICCV 2025 · Vision-language and multimodal › Multimodal LLMs · approach: Instruction-tuned multimodal assistants · 4 citations (OpenAlex)

Read the paper

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 Semantic-Enhanced Feature Extractor (SEFE) (ablation)LIRA with InternLM2-1.8B · RefCOCO valcIoU: 77.2 -> 79.4Improved
add SEFE (ablation)LIRA with InternLM2.5-7B · VQAv2accuracy: 75.1 -> 80.0Improved
add Interleaved Local Visual Coupling (ILVC) (ablation)LIRA with InternLM2-1.8B · ChairSChairS (lower better): 32.2 -> 29.2Improved
add ILVC (ablation)LIRA with InternLM2.5-7B · ChairSChairS (lower better): 43.6 -> 38.8Improved
add ILVC (ablation)LIRA-8B · AttrEvalAcc1: 22.3 -> 25.7Improved
co-train with segmentation data in addition to comprehension dataLIRA-2B · MME/MMBench/POPEcomprehension accuracy: -0.2% averageNo clear effect
co-train with segmentation data in addition to comprehension dataOMG-LLaVA · MMBenchaccuracy: 67.5 -> 47.9Hurt
use ILVC at inference when trained with two prompts on COCO instance segmentation (ablation)LIRA · COCO instance segmentationmIoU: 60.6 -> 58.9Hurt
include ILVC during training for instance segmentation (ablation)LIRA · COCO instance segmentationmIoU: 60.0 -> 60.6Improved
replace InternLM2-1.8B backbone with Qwen2-1.5B (ablation)LIRA · RefSeg averageaverage cIoU: 75.5 -> 76.7Improved

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