Seamless Human Motion Composition with Blended Positional Encodings

forward kick", 2.5s) ("walk slowly", 3.2s) ("get down on ground", 3s) ("crawl", 3.3s) ("walk", 2s) ("walk", 2s) ("walk", 2s) ("walk", 2s) ("walk", 2s) ...

Figure 1. We present FlowMDM, a diffusion-based approach capable of generating seamlessly continuous sequences of human motion from textual descriptions (left). The whole sequence is generated simultaneously and it does not require any postprocessing. FlowMDM also makes strides in the challenging problem of extrapolating and controlling periodic motion such as walking, jumping, or waving (right).