{"rewrite":{"id":"r_1175f8ece8cd0d08314ebcfe","clusterId":"c_37644a654535295860d7bad8","slug":"asilla-s-custom-vlm-boosts-security-detection-accuracy-by-80","model":"deepseek-v4-flash","headline":"Asilla's Custom VLM Boosts Security Detection Accuracy by 80%","summary":"Asilla has integrated its proprietary Vision-Language Model AsillaVision into its AI Security asilla system, claiming a detection accuracy improvement of up to 80% in pilot deployments. The lightweight 4B-parameter model runs on edge devices, combining with existing action recognition AI to distinguish between falls, fights, and other events. Full-scale deployment at large Japanese facilities began in June 2026.","whyItMatters":"Asilla's move from motion detection to contextual understanding with a custom VLM marks a practical shift in AI security, leveraging proprietary data and edge computing to address real-world facility needs.","webCardHtml":"\u003cp\u003eAsilla, founded in 2015, has been developing action recognition AI for security. Its AI Security asilla system, launched in 2022, is now deployed in over 200 facilities and has accumulated more than 8 million surveillance video data points. The company has now integrated a custom Vision-Language Model, AsillaVision, into the system. The 4B-parameter model runs on edge devices, combining with existing pose-estimation AI to understand context-distinguishing a fall from a person leaning against a wall. In pilot tests, detection accuracy for falls and fights improved by up to 80%. Full-scale deployment at large facilities began in June 2026, with plans to expand detection categories and add predictive features.\u003c/p\u003e","blueskyPost":"Asilla has integrated its custom VLM AsillaVision into its AI security system, claiming up to 80% better detection accuracy. The lightweight model runs on edge devices, combining with action recognition to understand context. Full deployment at large Japanese facilities started June 2026.","twitterPost":"Asilla integrates custom VLM AsillaVision into its AI security system, claiming up to 80% accuracy improvement. The edge-running model combines with action recognition to understand context. Full deployment at large Japanese facilities started June 2026.","threadsPost":null,"newsletterBlurb":"Asilla has integrated its proprietary Vision-Language Model AsillaVision into its AI Security asilla system, reporting a detection accuracy improvement of up to 80% in pilot deployments. The lightweight model runs on edge devices and combines with existing action recognition AI to distinguish between events like falls and fights. Full-scale deployment at large Japanese facilities began in June 2026, with plans to expand detection categories.","attributionJson":"[{\"source\":\"ASCII.jp\",\"url\":\"https://ascii.jp/elem/000/004/413/4413091/?rss\",\"title\":\"映像解析AIのアジラ、独自VLM「AsillaVision」を国内大型施設で本格展開へ\"}]","lintFlagsJson":null,"lintHits":0,"costUsd":0,"inputTokens":5695,"outputTokens":2512,"status":"published","repairAttempts":0,"nextRepairAt":null,"factsAttemptedAt":1782272664,"createdAt":"2026-06-24T03:38:15.000Z","publishedAt":"2026-06-24T03:40:24.000Z","updatedAt":"2026-06-24T03:40:24.000Z"},"cluster":{"id":"c_37644a654535295860d7bad8","canonicalTitle":"映像解析AIのアジラ、独自VLM「AsillaVision」を国内大型施設で本格展開へ","representativeArticleId":"a_32e5d3659eb1c1e1d5d65f58","sourceCount":1,"writtenSourceCount":1,"writeAttempts":0,"isSolo":true,"entitiesJson":"{\"anime_titles\":[],\"manga_titles\":[],\"work_titles\":[],\"studios\":[],\"people\":[],\"type\":\"news\",\"domain\":\"other\",\"is_roundup\":false}","contentType":"news","status":"published","firstSeenAt":"2026-06-24T01:00:00.000Z","lastSeenAt":"2026-06-24T01:00:00.000Z","updatedAt":"2026-06-24T03:40:24.000Z"},"attribution":[{"source":"ASCII.jp","url":"https://ascii.jp/elem/000/004/413/4413091/?rss","title":"映像解析AIのアジラ、独自VLM「AsillaVision」を国内大型施設で本格展開へ"}],"entities":{"anime_titles":[],"manga_titles":[],"work_titles":[],"studios":[],"people":[],"type":"news","domain":"other","is_roundup":false},"keyFacts":["Asilla integrated its custom Vision-Language Model AsillaVision into its AI Security asilla system, improving detection accuracy for falls and fights by up to 80% in pilot tests.","The 4B-parameter AsillaVision model runs on edge devices and combines with existing pose-estimation AI to distinguish between events like falls and fights.","Full-scale deployment of the updated system at large Japanese facilities began in June 2026.","Asilla's AI Security asilla system, launched in 2022, is deployed in over 200 facilities and has accumulated more than 8 million surveillance video data points."]}