In recent years,large quantities of effective and accurate algorithms on face alignment have been proposed,among which many of them use the mean shape as their initialized shapes and then further predict the final shapes of test images with different methods.These algorithms,however,are in lack of great robustness if there are big differences in facial expressions,head poses and illumination.This paper presents a Local-Shape-Based Combined(LSBC)model to creating a new face shape which is more accurate than ...
| 
推荐 
点击下列相关研究机构和相关文献作者，可以直接查到这些机构和作者被《中国知识资源总库》收录的其它文献，使您全面了解该机构和该作者的研究动态和历史。
