Improvements of 3D Object Detection with Part-based Models
基于部件的三维目标检测算法新进展

3D Object detection in 3D views is important and fundamental in the computer vision field. Its core issues include coping with the problems of partial occlusion, intra-class variance, cluttered background and multiple views. Most recently, with the development of part-based model, the occlusion and intra-class variance problems have been partly solved. Therefore, researchers have focused their attention mainly on multi-view problem. This paper gives detailed analysis of the 3D object detection problem with emphasis on the recent works related to this problem. At last, we summarize the algorithms introduced and give their comparisons.