Skip to content

cycoslee/NV_DeepstreamSDK_Excercise

Repository files navigation

NV_DeepstreamSDK_Excercise


This repository includes DeepstreamSDK hands-on material to understand how can develop IVA application with Gstreamer with plugins that DeepstreamSDK serves. Specially, this hands-on is working with Jetson Series but prepared docker container is for the Jetson NX. (TODO) It will be updated to have Dockerfile that could build with custom configurations.

Prerequisites

1. Host machine

  • SSH terminal
  • VLC player
  • (For TLT exercise) GPU machine is required & docker runtime environment
    • $ docker run --gpus all -it -v $(pwd):/workspace –w /usr/local/src -p 8888:8888 cycoslee/nv-deepstreamsdk-handson:tlt_host_210127

2. Jetson NX (based on Jetpack 4.4.1)

  • docker container pull
    • $ sudo docker run -ti --runtime=nvidia --rm --net=host -e DISPLAY=$DISPLAY -w /opt/nvidia/deepstream/deepstream-5.0 --device /dev/video0 -v /tmp/.X11-unix/:/tmp/.X11-unix -v $(pwd):/workspace cycoslee/nv-deepstreamsdk-handson:ds_nx_210127
  • Assets
    • IP camera(supports RTSP)
    • USB camera(webcam)

Exmaples

(Please check README file in the every each examples.)

  • 1_test_ds_input_video
  • 2_test_ds_input_webcam
  • 3_test_ds_input_uri
  • 4_test_ds_output_video
  • 5_test_ds_output_rtsp
  • 6_test_ds_output_rtsp_multi
  • 7_test_trt_benchmark
  • 8_test_ds_pose_estimation
  • 9_test_ds_yolov4
  • 10_test_ds_tlt_facemask