Research map / Vision-language and multimodal / Multimodal LLMs
LMM4-IC4K: A Large Multimodal Model Powered Integrated Circuit Footprint Geometry Understanding
Yida Wang, Taiting Lu, Runze Liu, Lanqing Yang, Zhe Chen, Yuehai Wang, Yixin Liu, Kaiyuan Lin, XiaoMeng Chen, Dian Ding, Yijie Li, Yifan Yang and 3 more
Read the paperCheck your idea against 137,217 papers
AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Introduces LMM4-IC4K, a large multimodal model-based approach and benchmark for understanding integrated circuit PCB footprint geometry from drawings.
Problem
Automated PCB footprint geometry labeling of integrated circuits is hard because footprint drawings are unstructured and abstractly annotated, and existing large multimodal models struggle with precise visual perception of these diagrams.
Approach
Adapts a large multimodal model to parse IC footprint drawings and label package geometry, with a dedicated IC footprint dataset (name suggests about 4K samples). Only the title and a truncated abstract were available, so the specific technique is not confirmed.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| train all three tasks in one chain-of-thought conversation rather than separate conversations (ablation) | LMM4-IC4K · ICGeoQA | overall IoU_IC (%): 63.5 -> 71.6 | Improved |
| train tasks 1 then 2 and 3 in one conversation vs all three chained (ablation) | LMM4-IC4K · ICGeoQA | overall IoU_IC (%): 62.6 -> 71.6 | Improved |
| train in three separate single-task rounds vs one chained conversation (ablation) | LMM4-IC4K · ICGeoQA | overall IoU_IC (%): 25.4 -> 71.6 | Improved |
| add synthetic pre-training stage before real-data fine-tuning (vs real only) (ablation) | LMM4-IC4K · ICGeoQA | overall IoU_IC (%): 65.1 -> 71.6 | Improved |
| two-stage synthetic then real vs mixing synthetic and real together (ablation) | LMM4-IC4K · ICGeoQA | overall IoU_IC (%): 68.2 -> 71.6 | Improved |
| synthetic first then real vs real first then synthetic (ablation) | LMM4-IC4K · ICGeoQA | overall IoU_IC (%): 24.7 -> 71.6 | Improved |
| add synthetic-only pre-training to base Qwen2-VL-7B (ablation) | Qwen2-VL-7B · ICGeoQA | overall IoU_IC (%): 1.7 -> 42.9 | Improved |
| increase real-world data from 0% to 100% (ablation) | LMM4-IC4K · ICGeoQA | overall IoU_IC (%): 42.9 -> 71.6 | Improved |
| add 4-shot domain examples to GPT-5 | GPT-5 · ICGeoQA | overall IoU_IC (%): 40.6 -> 43.0 | Improved |
| add 4-shot domain examples to Gemini 2.5 Flash | Gemini 2.5 Flash · ICGeoQA | overall IoU_IC (%): 16.5 -> 38.0 | Improved |
| add 4-shot domain examples to Qwen2-VL-7B | Qwen2-VL-7B · ICGeoQA | overall IoU_IC (%): 1.7 -> 3.8 | Improved |
| fine-tune with expert data instead of 4-shot prompting of GPT-5 | GPT-5 · ICGeoQA | overall IoU_IC (%): 43.0 -> 71.6 | Improved |
Limitations
- Only the abstract opening was available; full method and results could not be verified.
Closest papers by meaning
- ArchCAD-400K: A Large-Scale CAD drawings Dataset and New Baseline for Panoptic Symbol Spotting (NeurIPS 2025 · similarity 0.82)
- Functional Matching of Logic Subgraphs: Beyond Structural Isomorphism (NeurIPS 2025 · similarity 0.81)
- Point or Line? Using Line-based Representation for Panoptic Symbol Spotting in CAD Drawings (NeurIPS 2025 · similarity 0.81)
- DeepLayout: Learning Neural Representations of Circuit Placement Layout (ICML 2025 · similarity 0.81)
- CAD-GPT: Synthesising CAD Construction Sequence with Spatial Reasoning-Enhanced Multimodal LLMs (AAAI 2025 · similarity 0.80)
- GeoCAD: Local Geometry-Controllable CAD Generation with Large Language Models (NeurIPS 2025 · similarity 0.80)
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.
