no code implementations • 5 Dec 2022 • Zhangjian Ji, Zilong Wang, Ming Zhang, Yapeng Chen, Yuhua Qian
Recently, human pose estimation mainly focuses on how to design a more effective and better deep network structure as human features extractor, and most designed feature extraction networks only introduce the position of each anatomical keypoint to guide their training process.
no code implementations • 7 May 2022 • Zhangjian Ji, Kai Feng, Yuhua Qian, Jiye Liang
Another approaches can alleviate the temporal degeneration of learned filters by introducing the temporal regularizer, which depends on the assumption that the filers between consecutive frames should be coherent.
no code implementations • 25 May 2018 • Zhangjian Ji, Kai Feng, Yuhua Qian
In addition, to mitigate the issue of drift away from object further, we introduce inter-frame consistencies of local parts into our model.