AI Powered Cloud Vision Mixing for the online event industry

大型在线虚拟活动和网络研讨会已经成为向广大受众展示和交流信息的关键方面，需求在Covid大流行的推动下呈爆炸式增长。这类活动包括来自演讲者网络摄像头的信息流、视频流、演示幻灯片和小组讨论以及各种与会者互动和活动。由训练有素的制作工程师在幕后制作活动，确保所有与会者收到一个界面布局，使他们能够在任何给定的时间理解正在发生的事情。该项目寻求使用人工智能自动执行视觉混合过程，以快速准确地选择最合适的现场布局组合，消除对每个虚拟活动的人工制作人员的需要(这是一个巨大的增长障碍)，并降低对我们客户的服务成本。

Large scale online virtual events and webinars have become a critical aspect of presenting and communication information to large audiences, with explosive growth in demand driven by the Covid pandemic.Such events comprise a constantly varying mix of information streams from presenter webcams, video streams, presentation slides and panel discussions, as well as various attendee interactions and activities. Behind the scenes event production by trained production engineers ensures all attendees receive an interface layout that allows the to make sense of what is going on at any given time.This project seeks to automate the vision mixing process using artificial intelligence to quickly and accurately select the most appropriate layout mix live, removing the need for a human production crew for every virtual event (which is a huge growth barrier) and reducing the cost of service to our customers.