search

zhengthomastang / 2018AICity_TeamUW

The winning method in Track 1 and Track 3 at the 2nd AI City Challenge Workshop in CVPR 2018 - Official Implementation
http://openaccess.thecvf.com/content_cvpr_2018_workshops/w3/html/Tang_Single-Camera_and_Inter-Camera_CVPR_2018_paper.html
550 stars 198 forks source link

one difference between the provided code and the paper #30

Closed TuringKi closed 2 years ago

TuringKi commented 2 years ago

in your paper[1], you said:

The proposed camera self-calibration framework mainly depends on reliable human body segmentation and EDA to search for optimal locations of vanishing points and optimize the camera parameters, so that we can exploit the availability of human tracking and segmentation data for robust calibration

However, in this repository, i only found CAM_CAL, which used manually set or the default vanishing points.

May i ask about the implementation details about how to find the vanishing points?

1.Tang, Zheng, et al. "Multiple-kernel based vehicle tracking using 3D deformable model and camera self-calibration." arXiv preprint arXiv:1708.06831 (2017).

zhengthomastang commented 2 years ago

We didn't publish the code for that paper. You can try our latest calibration method here: https://github.com/zhengthomastang/Cal_PnP