<?xml version="1.0" encoding="UTF-8"?><oembed><type>video</type><version>1.0</version><html>&lt;iframe src=&quot;https://www.loom.com/embed/86c7dbac0dcb41af84bbbed72caa1acc&quot; frameborder=&quot;0&quot; width=&quot;1658&quot; height=&quot;1243&quot; webkitallowfullscreen mozallowfullscreen allowfullscreen&gt;&lt;/iframe&gt;</html><height>1243</height><width>1658</width><provider_name>Loom</provider_name><provider_url>https://www.loom.com</provider_url><thumbnail_height>1243</thumbnail_height><thumbnail_width>1658</thumbnail_width><thumbnail_url>https://cdn.loom.com/sessions/thumbnails/86c7dbac0dcb41af84bbbed72caa1acc-159fe3a9959f0c29.gif</thumbnail_url><duration>106.389</duration><title>FightLens Explains Boxing Moments in Real Time</title><description>This Loom explains FightLens, an AI web app that provides real-time explanations and win probability estimates for UFC and boxing fights. Max describes using YOLO to locate fighters, zooming in when fighters move closer or show attack signs, then generating live captions based on the action. VAST is used to bring video data and live events together to give the prediction model more confidence, and CloudFlash on CloudFly is used for fast updates with very little delay. The team plans to train and test the system on test flights to improve accuracy and reliability.</description></oembed>