Fast multi-frame motion estimation for H.264 and its applications to complexity-aware streaming

JVT/H.264 achieves higher compression efficiency than previous video coding standards such as MPEG4 and H.263. However, this comes at the cost of increased complexity due to the use of variable block-size, multiple reference frame motion compensation, and other advanced coding algorithms. In this paper, we present a fast motion estimation algorithm. This algorithm is based on an adaptive search strategy and a flexible multi-frame search scheme. Compared to the H.264 reference software JM 8.5, this algorithm achieves, on average, a 522% reduction of encoding time, with negligible PSNR drop and bit-rate increase. Performance comparison results are described, and the architecture of a complexity-aware streaming server that applies the algorithm to provide differentiated services is discussed.