Start Time: 16:20
End Time: 16:50
Traditional techniques to measure television viewership have relied on active engagement from the audience in the form of button clicks and surveys. Not only do these techniques bias the captured data, they also fail to measure viewers’ attention levels as they cannot deduce whether viewers are actually in front of the TV, and if so, whether they are actually watching. This talk describes how TVision is leveraging the latest advances in deep learning and computer vision to measure viewer attention on a per-second basis and categorize attentive reach by demographic, time of day, channel and program.