GitHub上的DeepStream参考应用程序

用例应用

*“ 360度端到端智能停车应用程序”。
感知+分析 <https://github.com/NVIDIA-AI-IOT/deepstream_360_d_smart_parking_application/tree/master/analytics_server_docker/kafka>
*面罩检测（TLT + DeepStream） <https://github.com/NVIDIA-AI-IOT/face-mask-detection>

*使用DeepStream进行修订<https://github.com/NVIDIA-AI-IOT/redaction_with_deepstream>_
*使用RetinaNet进行面部修饰<https://github.com/NVIDIA-AI-IOT/retinanet_for_redaction_with_deepstream>_
*使用DeepStream进行人数统计<https://github.com/NVIDIA-AI-IOT/deepstream-occupancy-analytics>

*DeepStream姿势估计<https://github.com/NVIDIA-AI-IOT/deepstream_pose_estimation>_
*DeepStream车牌检测和识别<https://github.com/NVIDIA-AI-IOT/deepstream_lpr_app>_

DeepStream的AI模型

*带有DeepStream的YoloV4 <https://github.com/NVIDIA-AI-IOT/yolov4_deepstream>

*使用Triton和DeepStream部署TensorFlow FasterRCNN模型<https://github.com/NVIDIA-AI-IOT/deepstream_triton_model_deploy>
*DeepStream中的TLT模型部署<https://github.com/NVIDIA-AI-IOT/deepstream_tlt_apps>

DeepStream功能样本

*使用DeepStream的背对背检测器<https://github.com/NVIDIA-AI-IOT/deepstream_reference_apps/tree/master/back-to-back-detectors> _
*使用DeepStream添加/删除运行时源 <https://github.com/NVIDIA-AI-IOT/deepstream_reference_apps/tree/master/runtime_source_add_delete>
*使用DeepStream使用NV Optical Flow进行异常检测<https://github.com/NVIDIA-AI-IOT/deepstream_reference_apps/tree/master/anomaly>_
*Python DeepStream应用程序（Python）中对SSD模型的自定义后处理<https://github.com/NVIDIA-AI-IOT/deepstream_python_apps/tree/master/apps/deepstream-ssd-parser>_
*从DeepStream管道（Python）中保存图像元数据<https://github.com/NVIDIA-AI-IOT/deepstream_python_apps/tree/master/apps/deepstream-imagedata-multistream> _

深度流ROS

*DeepStream应用程序的ROS2节点<https://github.com/NVIDIA-AI-IOT/ros2_deepstream> _

DeepStream5.0系列之yolov5使用
https://zongxp.blog.csdn.net/article/details/109444343