Meanwhile, speaker information can also be used in semantic video segmentation, high-level video features extraction, anchorperson detection and etc.

• 与此同时，说话人的信息还可以用于检测视频的语义切换、提取独白等和说话人有关的高层语义特征、新闻视频中的主持人查找等不同的视频处理任务。