Meng, Wenming
(Qingdao Research Institute, Beihang University,Qingdao,China)
,
Hu, Tao
(Qingdao Research Institute, Beihang University,Qingdao,China)
,
Shuai, Li
(Beihang University,State Key Laboratory of Virtual Reality Technology and Systems,Beijing,China)
In recent years, due to the deep learning and the development of computer vision, great progress has been made in the 3D human pose estimation from RGB images. However, due to the lack of depth information in RGB images, this task still faces great challenges. In this paper, we propose a method of a...
In recent years, due to the deep learning and the development of computer vision, great progress has been made in the 3D human pose estimation from RGB images. However, due to the lack of depth information in RGB images, this task still faces great challenges. In this paper, we propose a method of adversarial learning to estimate the 3D pose of the human body. Our framework consists of two parts, a pose generator and a discriminator. Using the 3D pose descriptor, we designed for adversarial learning can effectively increase the accuracy and visual effect of 3D pose estimation results. We performed ablation experiments on the public dataset, which is a good improvement compared to our baseline.
In recent years, due to the deep learning and the development of computer vision, great progress has been made in the 3D human pose estimation from RGB images. However, due to the lack of depth information in RGB images, this task still faces great challenges. In this paper, we propose a method of adversarial learning to estimate the 3D pose of the human body. Our framework consists of two parts, a pose generator and a discriminator. Using the 3D pose descriptor, we designed for adversarial learning can effectively increase the accuracy and visual effect of 3D pose estimation results. We performed ablation experiments on the public dataset, which is a good improvement compared to our baseline.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.