harm belt

Presentation abstract

Lifelike Communication – front-end audio and video technologies (Break-out sessie Immersive Experiences)

Communication is a basic human need. In our modern society however, families have split schedules or are living apart, and struggle to find the right balance between work, other activities, and relationships. They seek quality time together and opportunities for rich interactions. They want to keep in touch through tele-communication technology, and they want a more lifelike interaction. For tele healthcare and remote patient monitoring applications, lifelike interaction between the physician, the patient, and potentially a family member, is something that is highly desired.

Lifelike communication implies that the people using the technology enjoy full freedom in their movements, and the communication technology itself is hidden. It also implies large displays in order to be able to display remote people at sufficient size. Consequently, with such systems the distance between the people and the audio/video sensors and actuators is large. On top of that, the environment in which the technology is used can be severe in the sense that the acoustic and light conditions are bad. This often leads to bad audio and video quality, while for lifelike communication a high quality is required. Another important aspect in video communication is eye contact, which is not automatically preserved due to the misalignment between camera and display.

At Philips Research, over the past two decades much work was done on audio and video signal processing improving the quality of the microphone and camera signals before transmission and the loudspeaker and display signals after receiving. In this talk we address many of these technologies and provide several audio and video demonstration examples, enabling lifelike communication in the living room with the TV, and lifelike communication for healthcare. We present solutions for the communication of spatial sound and 3D video, the latter being important also for eye contact.


Harm Belt is principal research scientist with Philips Research, where since 1997 he has been working on various speech and 2D/3D video signal processing topics. Currently he is project leader in the Video and Image Processing group.

Leave a comment

Comments are closed.