Bell Labs has developed a video-conferencing system that enables video streams to be adjusted to user emotions so as to help maintain audience attention.

Video-Conferencing: Assessing Audience Emotion to Help Retain Attention

During a lecture it can sometimes prove a real struggle to keep the audience’s attention. This is all the more so when it comes to a video-conference session where image quality may not be up to standard for the viewers. Now a team of researchers at Alcatel-Lucent’s Bell Labs has been working on ways of optimising video streams in order to retain the attention of everyone taking part in a video-conference. The system that was presented on 20 June at the Bell Labs Open Days held at their Research Center at Villarceaux, near Paris, uses ‘affective computing’, i.e. devices and systems that can recognise, interpret and process human emotions, and applies this information to video-conferencing technology.

User-centric technology

The system uses two types of device for analysing user behaviour in order to suit the video stream optimally to an individual viewer. The first is a camera which uses facial recognition to pick up physical behaviour. The second is a sensor which measures brain waves and the user’s state of emotion or stress. “Measuring brain waves is the most avant-garde aspect of our method,” the researchers pointed out, underlining that “the system provides fast, efficient feedback on user attention.” Software based on a ‘hidden Markov model’ then picks up data stored on a server and aggregates it in order to determine the video stream which would best attract and hold user attention. In addition, the person delivering the video-conference can also intervene and generate video streams by using key words and key gestures.

Towards a new approach to visual streaming

The Bell Labs researchers further explained that: “This optimisation means that you can relay images which are more in line with what the users require or expect. The system has a number of potential fields of application.” The education field was one of the examples they gave. Online courses using video content could be among the first beneficiaries of the system, as information coming from the camera and the brain wave sensor will provide the video-conference lecturer with instant feedback so that s/he can relay video content which is more suited to the students’ concentration levels. The system might also be used for sports event broadcasts, enabling the viewer to enjoy more time viewing what s/he finds most interesting.

By Guillaume Parodi