This may not be run in your environment other than our teammates.

1. https://github.com/NVIDIA-AI-IOT/trt_pose Follow the example int the website above and copy all directories and files in tasks folder to tensorRT-yolo/tensorrt_demos (in your case run test.py)

2. run this command in tensorrt_demos

... 
$ cd tensorrt_demos git clone https://github.com/sugamza95/2020-JETSON-NX-TERM-PRJ.git

3. copy yolov4-tiny-416.cfg, yolov4-tiny-416.weights, yolov4-tiny-416.onnx, yolov4-tiny-416.trt in yolo of my github to your tensorrt_demos/yolo directory

$ cd 2020-JETSON-NX-TERM-PRJ 
$ cd yolo 
$ cp yolov4-tiny-416.cfg ../../yolo 
$ cp yolov4-tiny-416.weights ../../yolo 
......

4.copy trt_yolo_with_centernet.py to tensorrt_demos directory

$ cp trt_yolo_with_centernet.py .. 

5. cp yolo_with_plugins.py in the utils folder in my project directory to your /tensorrt_demos/utils folder

$ cd utils 
$ cp yolo_with_plugins.py ../../ 

5-1(optional). copy resnet18_baseline_att_224x224_A_epoch_249.py, resnet18_baseline_att_224x224_A_epoch_249_trt.py to tensorrt_demos

$ cd .. 
$ cp resnet18_baseline_att_224x224_A_epoch_249.py .. 
$ cp resnet18_baseline_att_224x224_A_epoch_249_trt.py .. 

6.change directories in trt_yolo_with_centernet.py ex)

2020-JETSON-NX-TERM-PRJ/trt_yolo_with_centernet.py

Line 201 in 3766936

out_video = cv2.VideoWriter('/home/ee201511281/street_2_term_prj.mp4', fourcc, 30.0 , (1280, 720))

7.run this command

$ python3 trt_yolo_with_centernet.py --video <video file> -m yolov4-tiny-416 

If you want to use your custom weights, cfg files refer to this github https://github.com/jkjung-avt/tensorrt_demos