EmoScribe Camera: A Virtual Camera System to Enliven Online Conferencing with Automatically Generated Emotional Text Captions

Abstract

Ideally, for lively discussions to occur during online meetings, the participants should turn on both their camera and microphone. In practice this is not always possible, and meeting participants may opt to use a text chat to communicate their ideas and reactions instead. However, text messages are also time-consuming and labor-intensive to type as well as omit many of the emotional cues available through visual and audio channels. To address these issues, we propose EmoScribe Camera, a virtual camera system that generates images of automatic text captions in real time and outputs them as a software-based virtual camera that simulates a physical camera. We report on the results of a user study evaluating the efficacy of EmoScribe Camera as an alternative communication channel during online conferences when participants have their camera and microphone turned off.