3D Human Motion Prediction Based Deep Learning
Nivedita Wani, Shailesh Pramod Bendale · International Journal of Advanced Research in Science Communication and Technology · 2022
A growing number of established methods utilising deep neural networks to handle the problem of human motion prediction have been developed as a result of deep learning's success in a wide range of computer vision and computer graphics jobs. Modern motion prediction techniques concentrate on resolving numerous problems to forecast accurately and regular human movement across time. We give a thorough analysis of deep learning-based human motion prediction techniques in this paper. The goal of this work and the human motion prediction challenge are first defined. On the basis of our suggested classification, we then present pertinent background information and a thorough list of motion prediction techniques. Next, we give a thorough analysis of the traits frequently employed in the literature and describe the evaluation procedures. We concluded by providing a quantitative comparison.