Multiple description coding techniques are used to encode video content on multiple independent streams. Each stream provides a minimum level of quality and can be decoded independently or more than one stream can be combined to provide a higher quality video output. Multiple description coding can be used to deliver multimedia content over multiple independent paths which benefit from higher aggregated bandwidth. A particular application of such techniques is when there is a need for smoothing (from the point of view of playback distortion) possible handover operations when a user availing from a multimedia streaming session performs a handover between different networks.
The project aims at investigating various multiple description-coding algorithms and implementing and testing such an encoder. The tests will be performed using both objective video quality metrics and subjective perceptual testing using human subjects.
The project outcome should be a functional encoder and results of objective and subjective testing.