Wyze Labs

AI Video Captioning

The student team explored technical solution of video captioning for indoor home scenarios, video content understanding by bridging visual and language information and leveraging video data to generate captions to describe the semantic content in video recordings. The team worked to do in-depth research on academia-leading solutions on video captioning and apply state-of-the-art methods to build machine learning models to generate descriptions on Wyze device captured videos. The model and software developed will be used to create meaningful captions that help users monitor their daily events and identify solutions that need their attention.