MechVQA: Benchmarking and Enhancing Multimodal LLMs on Comprehensive Mechanical Drawing Understanding

Hugging Face Daily Papers Papers

Summary

This paper introduces MechVQA, a dataset with 3.3k high-density mechanical engineering drawings and 21k question-answer pairs, along with the MechVL model that outperforms existing baselines by 7.57 percentage points on the MechVQA total score, advancing multimodal LLM understanding of mechanical drawings.

Multimodal Large Language Models (MLLMs) have demonstrated significant achievements in general visual question answering (VQA) tasks. However, they remain brittle on mechanical engineering drawings, where high annotation density and weak domain knowledge, compounded by unreliable spatial relation reasoning under strict projection rules and geometric constraints, make decisive cues easy to miss and frequently lead to wrong answers. To bridge this gap, we introduce the first comprehensive mechanical drawing understanding dataset, MechVQA, created through a semi-automated construction and quality-control pipeline. MechVQA contains 3.3k high-density pictures with 21K question-answer pairs, spanning 10 different fine-grained tasks across three capability levels: Recognition, Reasoning, and Judging, providing a testbed to evaluate and improve MLLM understanding on real-world mechanical drawings. On top of MechVQA, we then develop the MechVL model through a multi-stage training paradigm, building a strong domain-specialized baseline. Extensive experimental results demonstrate that MechVL outperforms the strongest closed-source baseline by 7.57 percentage points on the MechVQA total score, significantly enhancing mechanical drawing understanding ability and providing a reusable foundation for deploying MLLMs in mechanical design and inspection scenarios.
Original Article
View Cached Full Text

Cached at: 06/05/26, 10:07 AM

Paper page - MechVQA: Benchmarking and Enhancing Multimodal LLMs on Comprehensive Mechanical Drawing Understanding

Source: https://huggingface.co/papers/2605.30794

Abstract

Mechanical engineering drawing understanding is improved through a specialized dataset and domain-specific model that outperforms existing baselines by leveraging multi-stage training and high-density visual question answering annotations.

Multimodal Large Language Models(MLLMs) have demonstrated significant achievements in generalvisual question answering(VQA) tasks. However, they remain brittle onmechanical engineering drawings, where high annotation density and weakdomain knowledge, compounded by unreliablespatial relation reasoningunder strictprojection rulesandgeometric constraints, make decisive cues easy to miss and frequently lead to wrong answers. To bridge this gap, we introduce the first comprehensive mechanical drawing understanding dataset, MechVQA, created through a semi-automated construction and quality-control pipeline. MechVQA contains 3.3k high-density pictures with 21K question-answer pairs, spanning 10 different fine-grained tasks across three capability levels: Recognition, Reasoning, and Judging, providing a testbed to evaluate and improve MLLM understanding on real-world mechanical drawings. On top of MechVQA, we then develop theMechVL modelthrough amulti-stage training paradigm, building a strong domain-specialized baseline. Extensive experimental results demonstrate that MechVL outperforms the strongest closed-source baseline by 7.57 percentage points on the MechVQA total score, significantly enhancing mechanical drawing understanding ability and providing a reusable foundation for deploying MLLMs in mechanical design and inspection scenarios.

View arXiv pageView PDFAdd to collection

Get this paper in your agent:

hf papers read 2605\.30794

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2605.30794 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2605.30794 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2605.30794 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

Do VLMs Reason Like Engineers? A Benchmark and a Stage-wise Evaluation

arXiv cs.AI

This paper introduces EngVQA, a multimodal benchmark for evaluating engineering reasoning in vision-language models, along with an 8-stage automatic evaluation framework that enables fine-grained analysis of reasoning failures. It reveals substantial limitations in current VLMs' engineering reasoning capabilities.

Benchmarking Visual State Tracking in Multimodal Video Understanding

Hugging Face Daily Papers

Introduces VSTAT, a benchmark for evaluating visual state tracking in multimodal large language models (MLLMs) using 834 clips and 1,500 questions. Current MLLMs perform poorly compared to humans, failing at visual perception rather than reasoning.

VAMPS: Visual-Assisted Mathematical Problem Solving Benchmark

arXiv cs.AI

VAMPS is a new benchmark of 1,168 multimodal bilingual math problems designed to evaluate whether LLMs can benefit from constructing and reasoning over graphs/visualizations. Key finding: direct analytical solving surprisingly outperforms tool-enabled visual solving even on problems where plotting is a natural strategy.