Open pangshengwei opened 3 years ago
I see that the multi-camera tracking algorithm is not provided by the DeepStream SDK but rather it has been implemented from scratch by hand. Are there any papers or references that validate and benchmarks the performance of this implementation? https://github.com/NVIDIA-AI-IOT/deepstream_360_d_smart_parking_application/blob/b9b1f9e9aa84e379c573063fc5622ef50d38898d/tracker/code/mctrack/mctracker.py#L1625
I have the same doubt,the readme mentions usage of inter camera clustering which is used to perform the tracking.It would be nice if other references were added that would expalin this concept in more details
I see that the multi-camera tracking algorithm is not provided by the DeepStream SDK but rather it has been implemented from scratch by hand. Are there any papers or references that validate and benchmarks the performance of this implementation? https://github.com/NVIDIA-AI-IOT/deepstream_360_d_smart_parking_application/blob/b9b1f9e9aa84e379c573063fc5622ef50d38898d/tracker/code/mctrack/mctracker.py#L1625