[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-a-new-benchmark-judges-ai-video-erasing-like-humans-do":10,"sections":45},{"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":34,"tags":35,"sources":40,"feedback":44,"feedback_at":22,"cost_usd":44,"total_tokens":44},6406,"a-new-benchmark-judges-ai-video-erasing-like-humans-do","A New Benchmark Judges AI Video Erasing Like Humans Do","Researchers built VOR-Bench, a dataset and scoring model that grades video object removal by how convincing it looks to people, not just pixel math.","A new benchmark asks a simple question: does that erased object in your video actually look gone, or just smudged?\n\nResearchers released VOR-Bench, which bundles three pieces. VORD is a paired-video dataset with edited clips and masks drawn from model-generated, tool-rendered, and camera-captured sources. rMPAF is a framework that combines image-based object removal with fine-tuned video generation models to automatically produce realistic, motion-coherent paired videos. VOR-MDSM is a vision-language-model-based scorer trained across three evaluation dimensions to match human judgment rather than arithmetic pixel comparisons. The team reports its scores correlate with human ratings above 0.9, and plans to release the full benchmark and documentation for reproducibility.\n\nThat matters because current video object removal is often graded by metrics that check whether an edit matches a reference frame, not whether the erased area still tracks moving shadows, reflections, and occlusion as the camera moves - the details that actually convince a viewer nothing was ever there. Pairing a mixed-source dataset with a perception-aligned scorer gives developers a more honest yardstick than synthetic-only benchmarks typically offer.\n\nThat 0.9-plus correlation is the paper's own reported number. Whether VOR-MDSM holds up once outside researchers start scoring their own models against it is the actual test.","[\"video object removal\",\"ai benchmarks\",\"computer vision\",\"generative video\"]","2026-09-16T04:00:00.000Z","2026-09-17T14:24:15.752Z","2026-09-17T14:24:27.637Z","published",null,[24,30],{"id":25,"reviewer":26,"round":27,"reason":28,"status":29},"editor-r1","editor",1,"Cut or rework the closing paragraph — the snarky claim that 'promised for release' has 'historically meant anything from next week to eventually, maybe' is unsupported editorializing tacked onto the ending rather than closing on the reported facts.","resolved",{"id":31,"reviewer":26,"round":32,"reason":33,"status":29},"editor-r2",2,"The closing paragraph still introduces unsubstantiated claims not found in the source — that VOR has become 'a staple feature in consumer photo and video apps' and that 'reviewers often can't tell what's been erased' — cut or attribute these before closing on the reported facts.","ai",[36,37,38,39],"video object removal","ai benchmarks","computer vision","generative video",[41],{"name":42,"url":43},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2609.16878",0,{"sections":46},[47,51,56,61,66,70,74,79,84,88,93,98,103,108],{"name":48,"slug":34,"count":49,"latest_published_at":50},"AI",3853,"2026-09-17T08:27:09.000Z",{"name":52,"slug":53,"count":54,"latest_published_at":55},"Security","security",648,"2026-09-17T04:00:00.000Z",{"name":57,"slug":58,"count":59,"latest_published_at":60},"Policy","policy",338,"2026-09-11T04:00:00.000Z",{"name":62,"slug":63,"count":64,"latest_published_at":65},"Deals","deals",179,"2026-06-29T20:02:07.000Z",{"name":67,"slug":68,"count":69,"latest_published_at":55},"Hardware","hardware",154,{"name":71,"slug":72,"count":73,"latest_published_at":55},"Science","science",114,{"name":75,"slug":76,"count":77,"latest_published_at":78},"Consumer Tech","consumer-tech",99,"2026-09-09T17:27:33.000Z",{"name":80,"slug":81,"count":82,"latest_published_at":83},"Software","software",75,"2026-09-10T20:41:21.000Z",{"name":85,"slug":86,"count":87,"latest_published_at":55},"Dev Tools","dev-tools",73,{"name":89,"slug":90,"count":91,"latest_published_at":92},"Startups","startups",55,"2026-09-09T23:14:29.000Z",{"name":94,"slug":95,"count":96,"latest_published_at":97},"Gaming","gaming",43,"2026-09-10T12:18:06.000Z",{"name":99,"slug":100,"count":101,"latest_published_at":102},"General","general",41,"2026-09-08T01:57:23.000Z",{"name":104,"slug":105,"count":106,"latest_published_at":107},"Reviews","reviews",20,"2026-06-24T12:00:01.000Z",{"name":109,"slug":110,"count":111,"latest_published_at":112},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]