Spatio-temporal Manifold Learning for Human Motions via Long-horizon Modeling

Data-driven modeling of human motions is ubiquitous in computer graphics and vision applications. Such problems can be approached by deep learning on a large amount data. However, existing methods can be sub-optimal for two reasons. First, skeletal information has not been fully utilized. Unlike images, it is difficult to define spatial proximity in skeletal motions in the way that deep networks can be applied for feature extraction. Second, motion is time-series data with strong multi-modal temporal correlations between frames. A frame could lead to different motions; on the other hand, long-range dependencies exist where a number of frames in the beginning correlate to a number of frames later. Ineffective temporal modeling would either under-estimate the multi-modality and variance. We propose a new deep network to tackle these challenges by creating a natural motion manifold that is versatile for many applications. The network has a new spatial component and is equipped with a new batch prediction model that predicts a large number of frames at once, such that long-term temporally-based objective functions can be employed to correctly learn the motion multi-modality and variances. We demonstrate that our system can create superior results comparing to existing work in multiple applications.