Scientific posters should be editable and print-ready—not just visually plausible.</p>\n<p>We introduce <strong>PosterMELD</strong>, a structure-first multi-agent system that turns scientific papers into native, editable PowerPoint posters. It fixes template capacity before writing, coordinates specialized content, layout, and visual agents, and uses deterministic gates together with VLM review to repair only the failed regions.</p>\n<p>Across 621 papers, PosterMELD achieves an <strong>81.3% Print-Ready Rate</strong>—<strong>3.4× P2P</strong> and <strong>5.2× PosterGen</strong>—while preserving controllable design diversity and native PPTX editability at only <strong>$0.38 per request</strong>.</p>\n<p>We open-source the complete generation pipeline, 24 structural templates, benchmark, and evaluation tools.</p>\n<p><strong>One paper, many valid posters.</strong></p>\n","updatedAt":"2026-08-05T06:51:06.627Z","author":{"_id":"6747f6c62a6fb16e880d4cf3","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/XlaPiX_DzuGW5rtC8BFKx.png","fullname":"Jie","name":"T-Jack","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.7351037263870239},"editors":["T-Jack"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/XlaPiX_DzuGW5rtC8BFKx.png"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2608.02218","authors":[{"_id":"6a717241ec5082b9f872ceee","user":{"_id":"6747f6c62a6fb16e880d4cf3","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/XlaPiX_DzuGW5rtC8BFKx.png","isPro":false,"fullname":"Jie","user":"T-Jack","type":"user","name":"T-Jack"},"name":"Haojie Hu","status":"claimed_verified","statusLastChangedAt":"2026-08-04T09:04:32.448Z","hidden":false},{"_id":"6a717241ec5082b9f872ceef","name":"Chenhao Dang","hidden":false},{"_id":"6a717241ec5082b9f872cef0","name":"Yaojia Liu","hidden":false},{"_id":"6a717241ec5082b9f872cef1","name":"Hengrui Kang","hidden":false},{"_id":"6a717241ec5082b9f872cef2","name":"Conghui He","hidden":false},{"_id":"6a717241ec5082b9f872cef3","name":"Weijia Li","hidden":false}],"mediaUrls":["https://cdn-uploads.huggingface.co/production/uploads/6747f6c62a6fb16e880d4cf3/a7WMb-xKdBeZSlzFktxMo.png"],"publishedAt":"2026-08-03T00:00:00.000Z","submittedOnDailyAt":"2026-08-05T00:00:00.000Z","title":"PosterMELD: Multi-Agent Paper-to-Poster Generation for Controllable Design Diversity with Editable Print-Ready Outputs","submittedOnDailyBy":{"_id":"6747f6c62a6fb16e880d4cf3","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/XlaPiX_DzuGW5rtC8BFKx.png","isPro":false,"fullname":"Jie","user":"T-Jack","type":"user","name":"T-Jack"},"summary":"Scientific poster construction compresses a long multimodal paper into a readable, editable canvas. Existing systems hide request-level failures by scoring only completed outputs; direct image generation is not element-editable, while coding-agent workflows are costly. PosterMELD is a template-conditioned multi-agent pipeline: capacity-aware slots guide writing before rendering, and deterministic gates plus vision-language model (VLM) review route failures to bounded repair. Each accepted request exports editable PowerPoint (PPTX) and Portable Network Graphics (PNG) artifacts; explicit design controls yield same-paper variants. Across 621 papers, Print-Ready Rate (PRR) counts requests passing geometric, readability, asset-integrity, and obvious-factual-error checks, with native editability reported separately. A frozen VLM assigns conditional Craftsmanship-Harmony-Expressiveness (CHE) scores to print-ready outputs. PosterMELD attains 81.3% PRR, 3.4 times P2P's rate and 5.2 times PosterGen's, and the highest conditional CHE among generated methods with multiple print-ready outputs. Native editability and explicit design controls are retained at a mean cost of USD 0.38 per request, 3.5% of Codex+Skill's. Code and resources are available at https://github.com/Shannon4Science/PosterMELD.","upvotes":0,"discussionId":"6a717241ec5082b9f872cef4","projectPage":"https://jackey0903.github.io/PosterMELD/","githubRepo":"https://github.com/Shannon4Science/PosterMELD","githubRepoAddedBy":"user","githubStars":1},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[],"acceptLanguages":["en"],"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2608/2608.02218.md","query":{}}">

Get this paper in your agent:

Don't have the latest CLI?

Cite arxiv.org/abs/2608.02218 in a model README.md to link it from this page.

Cite arxiv.org/abs/2608.02218 in a dataset README.md to link it from this page.

Cite arxiv.org/abs/2608.02218 in a Space README.md to link it from this page.

Add this paper to a collection to link it from this page.