[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-researchers-build-interpretable-ai-to-grade-acne-severity":10,"sections":34},{"siteName":4,"siteTagline":5,"publisherName":4,"contactEmail":6},"The Revision","Tech news, decoded.","editor@therevision.news",{"gaMeasurementId":8,"adsenseClientId":9},"G-ZW2MV82GYR","ca-pub-8533917693782264",{"article":11},{"id":12,"slug":13,"title":14,"dek":15,"body_md":16,"tags_json":17,"published_at":18,"created_at":19,"updated_at":20,"status":21,"review_note":22,"review_notes":23,"image_url":22,"persona_id":22,"persona_name":22,"section":24,"tags":25,"sources":29,"feedback":33,"feedback_at":22,"cost_usd":33,"total_tokens":33},8065,"researchers-build-interpretable-ai-to-grade-acne-severity","Researchers Build Interpretable AI to Grade Acne Severity","A new framework called CG-HAF pairs whole-face analysis with lesion counts for more transparent acne grading, but struggles across different standards.","A new AI framework grades acne severity by showing its work instead of just spitting out a number.\n\nResearchers built CG-HAF, a system that combines two kinds of evidence instead of blending them into one opaque score. It averages severity predictions from several image classifiers trained on whole-face photos, then adds structured lesion data - count, detection confidence, and area - pulled from a separate object detector. A lightweight classifier fuses those two streams into a final grade on standard acne severity scales. On a widely used benchmark, the approach beat baselines that rely on whole-face evidence alone, with the biggest gains on the most severe cases.\n\nAcne grading is notoriously subjective even among dermatologists, and black-box AI scores make that worse, not better, for anyone trying to trust or audit them. By keeping the whole-face and lesion-level evidence separate and legible, CG-HAF gives clinicians and skincare apps something closer to a paper trail than a verdict.\n\nThe catch: when the researchers tested it on a second dataset that uses a different grading standard, accuracy dropped, and the follow-up analysis pinned much of the blame on mismatched grading criteria rather than the model's eyesight - a reminder that interpretable doesn't automatically mean portable.","[\"ai\",\"computer-vision\",\"healthtech\",\"interpretability\"]","2026-09-28T04:00:00.000Z","2026-09-28T08:01:43.800Z","2026-09-28T08:01:50.823Z","published",null,[],"ai",[24,26,27,28],"computer-vision","healthtech","interpretability",[30],{"name":31,"url":32},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2609.31326",0,{"sections":35},[36,39,43,48,53,58,62,67,72,77,82,87,91,96],{"name":37,"slug":24,"count":38,"latest_published_at":18},"AI",4750,{"name":40,"slug":41,"count":42,"latest_published_at":18},"Security","security",759,{"name":44,"slug":45,"count":46,"latest_published_at":47},"Policy","policy",399,"2026-09-27T18:39:02.000Z",{"name":49,"slug":50,"count":51,"latest_published_at":52},"Deals","deals",261,"2026-09-27T15:30:35.000Z",{"name":54,"slug":55,"count":56,"latest_published_at":57},"Hardware","hardware",188,"2026-09-27T20:46:36.000Z",{"name":59,"slug":60,"count":61,"latest_published_at":18},"Science","science",151,{"name":63,"slug":64,"count":65,"latest_published_at":66},"Consumer Tech","consumer-tech",135,"2026-09-26T14:30:00.000Z",{"name":68,"slug":69,"count":70,"latest_published_at":71},"Software","software",91,"2026-09-25T20:55:00.000Z",{"name":73,"slug":74,"count":75,"latest_published_at":76},"Dev Tools","dev-tools",84,"2026-09-26T04:20:58.000Z",{"name":78,"slug":79,"count":80,"latest_published_at":81},"Startups","startups",76,"2026-09-25T18:33:59.000Z",{"name":83,"slug":84,"count":85,"latest_published_at":86},"Gaming","gaming",48,"2026-09-25T18:35:21.000Z",{"name":88,"slug":89,"count":85,"latest_published_at":90},"General","general","2026-09-26T17:02:42.000Z",{"name":92,"slug":93,"count":94,"latest_published_at":95},"Reviews","reviews",30,"2026-09-24T20:07:31.000Z",{"name":97,"slug":98,"count":99,"latest_published_at":100},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]