The problem Lime solves

Often content creators receive negative reviews after publishing their videos on online video sharing platforms. Sometimes, their videos remain viewed only for a short duration, after which the viewer is not really interested in the content or may get bored. The income revenue of all these content/video uploaders is highly dependent on the viewing duration of their videos. So, we develop a service where content creators can have their videos evaluated, using the attention detection system we have developed. We collect the video from our client and we have a few trusted individuals watch these videos. We record the faces of those viewers for the duration for which they watch the video. Their facial expressions, gazes and eye pupil movements are detected and analysed by our attention detection model for emotions and we create a dashboard with metrics plotted in charts, where the user can get a grasp of how good the video would perform on a global platform. Thus, he can make changes accordingly before publishing the video.

Challenges we ran into

We had some trouble while sharing all our individual files. We took a break and figured it out eventually. We managed to improve the attention detection model as well which was a challenging task.