This chapter discusses the needs for standard datasets in the articulated pose estimation and tracking communities. It describes the datasets that are currently available and the performance of state-of-the-art methods on them. We discuss issues of ground-truth collection and quality, complexity of appearance and poses, evaluation metrics and partitioning of data. We also discusses limitations of current datasets and possible directions in developing new datasets for future use.
CITATION STYLE
Andriluka, M., Sigal, L., & Black, M. J. (2011). Benchmark Datasets for Pose Estimation and Tracking. In Visual Analysis of Humans (pp. 253–275). Springer London. https://doi.org/10.1007/978-0-85729-997-0_13
Mendeley helps you to discover research relevant for your work.