retinanet_traffic_3D
retinanet_traffic_3D copied to clipboard
Implementation for paper: "Detection of 3D bounding boxes of vehicles using perspective transformation for accurate speed measurement"
can you please let me know how to test this repo on a video and estimate speed in visual form. like this picture 
Has anyone got the problem like this one? It took too long to utilize GPU and the output image is filled with random boxes, after only few frames, the program...
I commented lines 335-366 and uncommented lines 315-333, then ran dataset_utils/fromBrnoCompSpeed.py to produce the training data. NameError happens : name 'pkl_paths' is not defined.
Hi, I tried test.py and it performed well(output the video with 3D BBox). However, I can't find the visualized estimated speed for the detections, and I can't find the speed-related...
Hi, I am wondering what's the usage in track_detections function and test_dataset function in dataset_utils/test.py? And what is the difference between test_dataset function and test_video function. Thanks in advance?
Can you please tell me how did you evaluate your speed estimation algorithm with ground truth provided by BrnoSpeedDataset?