Multi-hypothesis prediction
A video encoder for encoding a video into a data stream using motion compensated prediction for inter predicted blocks, comprising a hypothesis number control, configured to control a number of prediction hypotheses of the inter predicted blocks within a predetermined portion of the video to meet a predetermined criterion.
1. Video decoder for decoding a video from a data stream using motion compensated prediction for inter predicted blocks, configured to
determine a number of hypotheses for a predetermined inter predicted block dependent on a size or geometry of the block,
wherein the number of hypotheses is determined by comparing an upper limit for the number N of hypotheses dependent on
(m*w*h)/((w+k x )*(h+k y )), wherein
m is a factor, w is a block width of the predetermined inter predicted block, h is a block height of the predetermined inter predicted block and k x and k y indicate the size of an interpolation filter used in the motion compensated prediction.
2. Video decoder according to claim 1 , configure so that the number of prediction hypotheses indicates a maximum number of the hypotheses for the inter predicted block ( 80 ).
3. Video decoder according to claim 1 , configured to insert into the data stream information on the number of prediction hypotheses for the predetermined inter predicted block.
4. Video decoder according to claim 3 , configured to, based on the size or the geometry of the predetermined inter predicted block, determine an allowed number interval, and, using a binarization of the allowed number interval, code the information into the data stream.
5. Video decoder according to claim 4 , configured so that a maximum of the allowed number interval is the lower, the smaller the size of the predetermined inter predicted block is, and the binarization of the allowed number interval comprises a mean codeword length which is the lower, the lower the maximum of the allowed number interval is.
6. Video decoder according to claim 1 , configured so that the number of hypotheses is the larger, the larger the predetermined inter predicted block ( 80 ) is.
7. Method for decoding a video from a data stream using motion compensated prediction for inter predicted blocks, comprising
determining a number of hypotheses for a predetermined inter predicted block dependent on a size or geometry of the block,
wherein the number of hypotheses is determined by comparing an upper limit for the number N of hypotheses dependent on
(m*w*h)/((w+k x )*(h+k y )), wherein
m is a factor, w is a block width of the predetermined inter predicted block, h is a block height of the predetermined inter predicted block and k x and k y indicate the size of an interpolation filter used in the motion compensated prediction.
8. A non-transitory digital storage medium having a computer program stored thereon to perform the method for decoding a video from a data stream using motion compensated prediction for inter predicted blocks, comprising
determining a number of hypotheses for a predetermined inter predicted block dependent on a size or geometry of the block,
wherein the number of hypotheses is determined by comparing an upper limit for the number N of hypotheses dependent on
(m*w*h)/((w+k x )*(h+k y )), wherein
m is a factor, w is a block width of the predetermined inter predicted block, h is a block height of the predetermined inter predicted block and k x and k y indicate the size of an interpolation filter used in the motion compensated prediction,
when said computer program is run by a computer.
9. Video encoder for encoding a video into a data stream using motion compensated prediction for inter predicted blocks, configured to
determine a number of hypotheses for a predetermined inter predicted block dependent on a size or geometry of the block,
wherein the number of hypotheses is determined by comparing an upper limit for the number N of hypotheses dependent on
(m*w*h)/((w+k x )*(h+k y )), wherein
m is a factor, w is a block width of the predetermined inter predicted block, h is a block height of the predetermined inter predicted block and k x and k y indicate the size of an interpolation filter used in the motion compensated prediction.