Multiscale motion-aware and spatial-temporal-channel contextual coding network for learned video compression

Yiming Wang, Qian Huang, Bin Tang, Xin Li, Xing Li

Abstract: Highlights•MS-MAM enhances spatial–temporal motion information in a coarse-to-fine manner.•STCCM leverages the spatial, temporal and channel contexts to reduce bitrate.•Uneven channel grouping scheme achieves a trade-off between the complexity and RD.•Our proposed MASTC-VC beats HM and VTM in PSNR and MS-SSIM metrics, respectively.