We present the content deformation field CoDeF as a new type of video
representation, which consists of a canonical content field aggregating the
static contents in the entire video and a temporal deformation field recording
the transformations from the canonical image (i.e., rendered