Video coding using a self-adaptive redundant dictionary consisting of spatial and temporal prediction candidates | IEEE Conference Publication | IEEE Xplore