STZhang/cs231a_project
This problem takes an input of an RGB-D image in the NYU-RGBD dataset. The goal is to perform 3D object recognition and indoor scene classification. These two tasks are tackled jointly in our holistic model, that is, some constraints are placed among scenes and these objects when reasoning about object detection, which helps boost the performance of scene classification and object detection.
正在读取项目信息…