Abstract by Courtni Byun
Investigating Video Annotation by Non-Experts
The collection and annotation of activity data has traditionally been an expensive and time-consuming process. The current collection process includes expert collectors simultaneously recording video and sensor data and then carefully coordinating the video and sensor data before expert annotators can begin the task of annotation. Our current research seeks to answer the questions of whether video data is in fact always necessary and whether non-experts can effectively both collect and annotate video and sensor data. We have created a Video Annotation Tool (VAT), which presents video and sensor data in a simple GUI. We have collected multiple forms of data with the intent of conducting a user study, using VAT, to determine the necessity of video data and the effectiveness of non-expert annotators.