Microsoft adds a bunch of analytics tools to Azure Media Services

Azure Media Services is a cloud platform for on demand as well as live video content delivery and management. In the annual National Association of Broadcaster’s show, it announced a slew of new features according to a report in TechCrunch. The individual announcements for each of these features have been going live on Microsoft’s Azure site for a while. Microsoft added a new Indexer, for recognising and tagging audio content in video. These can be used to generate close captions, transcripts or a searchable database. Eight languages are currently supported which include English, Spanish, German, Italian, French, Portuguese, Arabic and Chinese. The feature works for sources with unmixed audio without background noise such as interviews, meetings or lectures. Face detection can track faces across a video duration, assign individual tags and generate meta data of the location of any face in the frame at any point in time in the video. There is also an emotion detector, that can assign feelings to faces, including happiness, sadness, fear or anger. This can be used to gauge audience reactions in a conference or a keynote. The video summarisation feature rapidly generates a “video thumbnail” by pulling the dramatic moments from a lengthy video into a small clip. There is a configurable preset that allows users to tweak the values of the concatenation and shot detection that is used to generate the video thumbnail. Hyperlapse was actually  added to Azure back in May 2015. The feature produces smooth timelapses from unsteady camera footage. Aimed at video captured from drones, action cameras or hand held devices, the image stabilisation is used to generate short, attractive clips out of long and shaky footage. Other features include security during the  video management and delivery, as well as machine learning bots that can automate parts of the process.