版权所有:内蒙古大学图书馆 技术提供:维普资讯• 智图
内蒙古自治区呼和浩特市赛罕区大学西街235号 邮编: 010021
作者机构:Department of Computer Science ETH Zürich Max Planck Institute for Intelligent Systems Tübingen Germany
出 版 物:《arXiv》 (arXiv)
年 卷 期:2021年
核心收录:
摘 要:Frame rate greatly impacts the style and viewing experience of a video. Especially for footage that depicts fast motion, higher frame rate equates to smoother imagery and reduced motion blur artifacts. Upsampling of low-frame-rate videos of human activity is an interesting yet challenging task with many potential applications ranging from gaming to entertainment and sports broadcasting. The main difficulty in synthesizing video frames in this setting stems from the highly complex and non-linear nature of human motion and the complex appearance and texture of the body. We propose to address these issues in a motion-guided frame-upsampling framework that is capable of producing realistic human motion and appearance. A novel motion model is trained to interpolate the non-linear skeletal motion between frames by leveraging a large-scale motion-capture dataset (AMASS). The high-frame-rate pose predictions are then used by a neural rendering pipeline to produce the full-frame output, taking the pose and background consistency into consideration. Our pipeline only requires low-frame-rate videos and unpaired human motion data but does not require high-frame-rate videos for training. Furthermore, we contribute the first evaluation dataset that consists of high-quality and high-frame-rate videos of human activities for this task. Compared with state-of-the-art video interpolation techniques, our method produces interpolated frames with better quality and accuracy, which is evident by state-of-the-art results on pixel-level, distributional metrics and comparative user evaluations. Our code and the collected dataset are available at https://***/Render-In-Between. Copyright © 2021, The Authors. All rights reserved.