A Method for 3D Human Reconstruction Animation
A human body, three-dimensional technology, applied in the field of three-dimensional dynamic graphics processing, can solve the problems of insufficient human-computer interaction function, poor user experience, long modeling time, etc., to achieve good human-computer interaction effect, reduce preparation workload, and improve model accuracy. Effect
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0060] A method for a three-dimensional human body reconstruction animation involved in the present embodiment, the process steps are as follows:
[0061] S1. Obtain the 3D pose estimation of the human body
[0062] The video frame sequence V={I composed of 16 consecutive frames of images 1 ,I 2 ,…,I 16 }, using the HRNet (high-resolution network) model to extract high-resolution features from a single image, and obtain a high-resolution feature sequence F={f 1 ,f 2 ,…,f 16 }, the high-resolution feature sequence F={f 1 ,f 2 ,…,f 16 } Input the time generator to obtain the human pose parameter sequence P={p 1 ,p 2 ,…,p 16 }, where the human pose parameter p corresponding to the i-th frame image i =(x i ,X i ,θ i ,β i ), x i , X i , θ i , β i is the predicted value generated by the time generator, x iRepresents the 23 joint points of the human body in the i-th frame image B={b head ,b left_hand ,b right_elbow ,…,b neck } two-dimensional coordinates, its ...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


