Abstract: Highlights•Constructing a dataset of template actions used in scenarios of service industries.•Proposing a simple yet effective method for generating 2D virtual human videos.•Generating gestures and emotional lip-sync expressions using neural network.
Loading