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
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add Semantic-Enhanced Feature Extractor (SEFE) (ablation) | LIRA with InternLM2-1.8B · RefCOCO val | cIoU: 77.2 -> 79.4 | Improved |
| add SEFE (ablation) | LIRA with InternLM2.5-7B · VQAv2 | accuracy: 75.1 -> 80.0 | Improved |
| add Interleaved Local Visual Coupling (ILVC) (ablation) | LIRA with InternLM2-1.8B · ChairS | ChairS (lower better): 32.2 -> 29.2 | Improved |
| add ILVC (ablation) | LIRA with InternLM2.5-7B · ChairS | ChairS (lower better): 43.6 -> 38.8 | Improved |
| add ILVC (ablation) | LIRA-8B · AttrEval | Acc1: 22.3 -> 25.7 | Improved |
| co-train with segmentation data in addition to comprehension data | LIRA-2B · MME/MMBench/POPE | comprehension accuracy: -0.2% average | No clear effect |
| co-train with segmentation data in addition to comprehension data | OMG-LLaVA · MMBench | accuracy: 67.5 -> 47.9 | Hurt |
| use ILVC at inference when trained with two prompts on COCO instance segmentation (ablation) | LIRA · COCO instance segmentation | mIoU: 60.6 -> 58.9 | Hurt |
| include ILVC during training for instance segmentation (ablation) | LIRA · COCO instance segmentation | mIoU: 60.0 -> 60.6 | Improved |
| replace InternLM2-1.8B backbone with Qwen2-1.5B (ablation) | LIRA · RefSeg average | average cIoU: 75.5 -> 76.7 | Improved |
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.
- OMG-LLaVA-8B — OMG-LLaVA: Bridging Image-level, Object-level, Pixel-level Reasoning and Understanding (NeurIPS 2024) · better on 7 of 8 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
- PARTONOMY: Large Multimodal Models with Part-Level Visual Understanding (NeurIPS 2025 · similarity 0.87)
- OmniSegmentor: A Flexible Multi-Modal Learning Framework for Semantic Segmentation (NeurIPS 2025 · similarity 0.87)
- MMIE: Massive Multimodal Interleaved Comprehension Benchmark for Large Vision-Language Models (ICLR 2025 · similarity 0.86)
- Emergent Open-Vocabulary Semantic Segmentation from Off-the-shelf Vision-Language Models (CVPR 2024 · similarity 0.86)
- Perceive Anything: Recognize, Explain, Caption, and Segment Anything in Images and Videos (NeurIPS 2025 · similarity 0.86)
- MMIU: Multimodal Multi-image Understanding for Evaluating Large Vision-Language Models (ICLR 2025 · similarity 0.86)
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.
