Multi-Frame Motion-Compensated Prediction

Multi-frame motion-compensated prediction is a concept to increase the efficiency of motion-compensated prediction in hybrid video coding.

Multi-frame motion-compensated prediction extends the spatial displacement vector utilized in block-based hybrid video coding by a variable time delay permitting the use of more pictures than the previously decoded one for motion compensated prediction. This concept is illustrated in Figure 1.

Figure 2 shows rate-distortion curves that compare the coding efficiency of H.264/MPEG-4 AVC with different numbers of used reference pictures. As can be seen, increasing the number of used reference pictures generally increases the coding efficiency.

Multi-frame motion-compensated prediction in video coding standards

Multi-frame motion-compensated prediction is supported in the following video coding standards:

For H.264/MPEG-4 AVC and HEVC, the concept of multi-frame motion-compensated prediction was already included in the first Test Models and forms an integral part of the standards.

The concept of multi-frame motion-compensated prediction also builds that basis for a number of further developments for increasing coding efficiency or providing additional functionalities, including:

  • The generalized B picture concept in H.264/MPEG-4 AVC [4]
  • Multi-hypotheses motion-compensated prediction [3]
  • The concept of hierarchical prediction structures
  • The MVC extension of H.264/MPEG-4 AVC