This paper presents a novel adversarial deep neural network to estimate human poses from still images, such as those obtained from CCTV and the Internet-of-Things (IoT) devices. Specifically, the proposed adversarial deep neural network exhibits the spatial hierarchy of human body parts considering the fact that predicting the position of some parts is more challenging than others. The generative and the discriminative portions of the proposed adversarial deep neural network are designed to encode the spatial relationship between the parts in the first stage of the hierarchy (parents) and the parts in the second stage of the hierarchy (children). Each of the generator and the discriminator networks is designed as two components, which are sequentially connected together to infer rich appearance potentials and to encode not only the likelihood of the part’s existence but also the relationships between each body part and its parent. The method is evaluated on three different datasets, whose findings suggest that the proposed network achieves comparable results with other competing state-of-the-art approaches.
RADWAN, I. H. I., Moustafa, N., Keating, B., Choo, K-K. R., & GOECKE, R. (2019). Hierarchical Adversarial Network for Human Pose Estimation. IEEE Access, 7, 103619-103628. https://doi.org/10.1109/ACCESS.2019.2931050