Improving Learned Video Compression by Exploring Spatial Redundancy | IEEE Conference Publication | IEEE Xplore