Accepted by ICML2026</p>\n","updatedAt":"2026-06-05T07:49:13.943Z","author":{"_id":"642f6c64f945a8a5c9ee5b5d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/642f6c64f945a8a5c9ee5b5d/V_4S_39gZc3ttiO4rXccj.png","fullname":"XiaofengShi","name":"MonteXiaofeng","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":8,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.985776424407959},"editors":["MonteXiaofeng"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/642f6c64f945a8a5c9ee5b5d/V_4S_39gZc3ttiO4rXccj.png"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2605.30794","authors":[{"_id":"6a2231093490a593e87b1431","name":"Qian Kou","hidden":false},{"_id":"6a2231093490a593e87b1432","name":"Xiaofeng Shi","hidden":false},{"_id":"6a2231093490a593e87b1433","name":"Yulin Li","hidden":false},{"_id":"6a2231093490a593e87b1434","name":"Xiaosong Qiu","hidden":false},{"_id":"6a2231093490a593e87b1435","name":"Xinyang Wang","hidden":false},{"_id":"6a2231093490a593e87b1436","name":"Hua Zhou","hidden":false},{"_id":"6a2231093490a593e87b1437","name":"Cao Dongxing","hidden":false}],"mediaUrls":["https://cdn-uploads.huggingface.co/production/uploads/642f6c64f945a8a5c9ee5b5d/AgviRwM88j60cw2slP9tb.png"],"publishedAt":"2026-05-29T00:00:00.000Z","submittedOnDailyAt":"2026-06-05T00:00:00.000Z","title":"MechVQA: Benchmarking and Enhancing Multimodal LLMs on Comprehensive Mechanical Drawing Understanding","submittedOnDailyBy":{"_id":"642f6c64f945a8a5c9ee5b5d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/642f6c64f945a8a5c9ee5b5d/V_4S_39gZc3ttiO4rXccj.png","isPro":false,"fullname":"XiaofengShi","user":"MonteXiaofeng","type":"user","name":"MonteXiaofeng"},"summary":"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.","upvotes":1,"discussionId":"6a2231093490a593e87b1438","ai_summary":"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.","ai_keywords":["Multimodal Large Language Models","visual question answering","mechanical engineering drawings","domain knowledge","spatial relation reasoning","projection rules","geometric constraints","MechVQA dataset","MechVL model","multi-stage training paradigm"],"ai_summary_model":"Qwen/Qwen2.5-Coder-32B-Instruct","organization":{"_id":"61be9739d2f9358e24ca0a4f","name":"BAAI","fullname":"Beijing Academy of Artificial Intelligence","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/1664511063789-632c234f42c386ebd2710434.png"}},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"642f6c64f945a8a5c9ee5b5d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/642f6c64f945a8a5c9ee5b5d/V_4S_39gZc3ttiO4rXccj.png","isPro":false,"fullname":"XiaofengShi","user":"MonteXiaofeng","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"organization":{"_id":"61be9739d2f9358e24ca0a4f","name":"BAAI","fullname":"Beijing Academy of Artificial Intelligence","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/1664511063789-632c234f42c386ebd2710434.png"},"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2605/2605.30794.md"}">
MechVQA: Benchmarking and Enhancing Multimodal LLMs on Comprehensive Mechanical Drawing Understanding
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 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.
Community
Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.
Tap or paste here to upload images
Cite arxiv.org/abs/2605.30794 in a model README.md to link it from this page.
Cite arxiv.org/abs/2605.30794 in a dataset README.md to link it from this page.
Cite arxiv.org/abs/2605.30794 in a Space README.md to link it from this page.
Discussion (0)
Sign in to join the discussion. Free account, 30 seconds — email code or GitHub.
Sign in →No comments yet. Sign in and be the first to say something.