Video frame prediction with dual-stream deep network emphasizing motions and content details

Open Webpage

Qingming Huang, Zhongxiao Li, Liying Zheng, Tianyi Yang

Published: 01 Jan 2022, Last Modified: 13 Nov 2024Appl. Soft Comput. 2022EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•A motion decoder is added to MCnet to emphasize the motion information.•Multi-level content details are introduced to alleviate the ambiguousness of a prediction.•WGAN-GP framework is used to construct our model.•A composite loss is designed to guide our model to focus on motions and content details.