Abstract: Techniques for video encoding based on viewer feedback are described herein. Feedback data collected from viewers of video content may be used to determine a provider score value and a provide duration value. The provider score value may indicate a quantity of viewers for content from the provider. The provider duration value may indicate a transmission duration for content from the respective provider. An upgrade threshold may be determined, which is a value associated with assignment of additional channels for encoding of video content. When video content from the provider is available for transmission, the provider score for the provider may be compared to the upgrade threshold value to determine whether the additional channels will be assigned for encoding of the video content. In some examples, multiple upgrade thresholds may be employed, for example corresponding to multiple time ranges identified based on the provider duration value.

Abstract: An encoder for encoding a video signal comprising a sequence of pictures, comprising a motion estimator for determining a motion estimation among the sequence of pictures, the motion estimation enabling a motion-compensated prediction of a pre-determined set of pictures of the sequence of pictures to obtain predicted pictures and residual data representing a deviation between the predetermined set of pictures and the predicted pictures, a parameter setter for setting parameters of an optimization problem based on the motion estimation and the predetermined set of pictures, and a solver for solving the optimization problem with the parameters being set, by means of convex optimization in order to yield optimized picture data that represents an encoded version of the residual data, is described. An improvement of the trade-off between rate and distortion is achieved by incorporating the motion estimation into the construction or set-up of the optimisation problem to be solved.

Abstract: An encoder for encoding a video signal comprising a sequence of pictures, comprising a motion estimator for determining a motion estimation among the sequence of pictures, the motion estimation enabling a motion-compensated prediction of a pre-determined set of pictures of the sequence of pictures to obtain predicted pictures and residual data representing a deviation between the predetermined set of pictures and the predicted pictures, a parameter setter for setting parameters of an optimization problem based on the motion estimation and the predetermined set of pictures, and a solver for solving the optimization problem with the parameters being set, by means of convex optimization in order to yield optimized picture data that represents an encoded version of the residual data, is described. An improvement of the trade-off between rate and distortion is achieved by incorporating the motion estimation into the construction or set-up of the optimisation problem to be solved.