Amazon Rekognition Video segment detection now supports four new segment types, improves accuracy
Amazon Rekognition Video is a machine learning (ML) based service that can analyze videos to detect objects, people, faces, text, scenes, activities, and inappropriate content. Video segment detection makes it easy for media companies to streamline common operational tasks by automatically detecting frame-accurate silent black frames, end credits, color bars, and shot changes in video files. Starting today, customers can detect four additional segment types – opening credits, content segments, slates, and studio logos. In addition, customers get improved accuracy for shot detection and complex end credit sequences, along with new filtering options for black frame detection to handle varied file quality and color ranges.
Amazon Rekognition Video is a machine learning (ML) based service that can analyze videos to detect objects, people, faces, text, scenes, activities, and inappropriate content. Video segment detection makes it easy for media companies to streamline common operational tasks by automatically detecting frame-accurate silent black frames, end credits, color bars, and shot changes in video files. Starting today, customers can detect four additional segment types – opening credits, content segments, slates, and studio logos. In addition, customers get improved accuracy for shot detection and complex end credit sequences, along with new filtering options for black frame detection to handle varied file quality and color ranges.
Video segment detection enables operations and media supply chain teams to streamline highly manual and tedious content preparation and quality control (QC) tasks for video on demand (VOD) applications. Common tasks include: adding markers like ‘Next Episode’ and ‘Skip Intro’, detecting pre-inserted ad break markers indicated by silent black frames, and removing unwanted sections such as slates and color bars before publishing. With video segment detection, customers can reduce asset review time from hours to minutes, cut down the cost of manual operations by 2-5x, and prepare large volumes of archival or third-party content for streaming at short turnaround. This helps broadcasters and streaming platforms to free up operator resources for higher value work, create interactive skip prompts to keep viewers engaged, and simplify ad insertion workflows for easier monetization.
Video segment detection updates for Amazon Rekognition Video will be launched in waves. The update is already available in US East (Northern Virginia), US West (Oregon), Asia Pacific (Mumbai), Asia Pacific (Sydney), and Asia Pacific (Seoul) today. It will be available in Europe (Ireland), Europe (Frankfurt), Europe (London), Asia Pacific (Singapore), US West (Northern California), US East (Ohio), and Asia Pacific (Tokyo) on August 9th. To get started, please read our blog, refer to our documentation, and download the latest AWS SDK.