Beyond Pixels Leveraging Geometry and Shape Cues for Online[ICRA18]阅读笔记

Main contributions:

1. Device a pairwise costs for object tracking based on several 3D cues.
2. The costs are agnostic to the data association method.
3. Can incorporated into any optimization framework.

The efficacy of the monocular 3D cues:

1. 前两行是帧t与帧t+1以及它们各自的bounding boxes。

2. 通过将帧t中的对象lifting到3D并ballooning它们的位置来将其project到在t+1时刻观察到的图像上，在此映射区域寻找匹配对象并计算3D-2Dcost，大大地减少了搜索区域，降低了配对成本。

3. 通过将仅在此映射区域的检测backproject到3D并基于3Dvolume重叠计算3D-3Dcost（代码实现是通过3D凸包重叠）。

4. 混合各类cost(不仅仅是上述两个),使用匈牙利关联模式进行数据关联。

5. odometry estimates obtained from ORB-SLAM。

Composition of costs:

①中的五元组分别为检测的bounding box左上角的(x,y)坐标，bounding box的宽与高(w,h),以及bounding box中检测器的置信度。
公式(1)中的第一项是一个对象类别的平均形状，第二项中的V是表征平均形状形变方向的形变基础（一组特征向量）。

Illustration for understanding the concept of 3D-2D and 3D-3D costs:

Results

引用文献

（前面没缩进，见谅）
[7]　　J. K. Murthy, G. S. Krishna, F. Chhaya, and K. M. Krishna, “Reconstructing　vehicles from a single image: Shape priors for road scene understanding,” in Proceedings of the IEEE Conference on Robotics and Automation， 2017.
[8]　　J. K. Murthy, S. Sharma, and M. Krishna, “Shape priors for real-time monocular object localization in dynamic　environments,” in Proceedings of the IEEE Conference on Intelligent Robots and Systems(In Press), 2017.
[22] 　S. Song and M. Chandraker, “Joint sfm and detection cues for monocular 3d localization in road scenes,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2015.