Sample Configurations and Streams

Contents of the package

This section provides information about included sample configs and streams.

• samples: Directory containing sample configuration files, streams, and models to run the sample applications.

• samples/configs/deepstream-app: Configuration files for the reference application:

  • source30_1080p_dec_infer-resnet_tiled_display_int8.txt: Demonstrates 30 stream decodes with primary inferencing. (For dGPU and Jetson AGX Xavier platforms only.)

  • source30_1080p_dec_infer-resnet_tiled_display_int8.yml: YAML based config file to demonstrate 30 stream decode with primary inferencing. (For dGPU and Jetson AGX Xavier platforms only.)

  • source4_1080p_dec_infer-resnet_tiled_display_int8.txt: Demonstrates four stream decodes with primary inferencing, object tracking, and three different secondary classifiers. (For dGPU and Jetson AGX Xavier platforms only.)

  • source4_1080p_dec_infer-resnet_tracker_sgie_tiled_display_int8.yml: YAML based config file to demonstrate four stream decode with primary inferencing, object tracking, and three different secondary classifiers. (For dGPU and Jetson AGX Xavier platforms only.)

  • source4_1080p_dec_infer-resnet_tracker_sgie_tiled_display_int8_gpu1.txt: Demonstrates four stream decodes with primary inferencing, object tracking, and three different secondary classifiers on GPU 1 (for systems that have multiple GPU cards). For dGPU platforms only.

  • config_infer_primary.txt: Configures a nvinfer element as primary detector.

  • config_infer_primary.yml: YAML based config file to configure a nvinfer element as primary detector.

  • config_infer_secondary_carcolor.txt, config_infer_secondary_carmake.txt, config_infer_secondary_vehicletypes.txt: Configure a nvinfer element as secondary classifier.

  • config_infer_secondary_carcolor.yml,config_infer_secondary_carmake.yml, config_infer_secondary_vehicletypes.yml: YAML based config file to configure a nvinfer element as secondary classifier.

  • config_tracker_IOU.yml: Config file for IOU tracker.

  • config_tracker_NvSORT.yml: Config file for NvSORT tracker.

  • config_tracker_NvDeepSORT.yml: Config file for NvDeepSORT tracker.

  • config_tracker_NvDCF_accuracy.yml: Config file for NvDCF tracker for higher accuracy.

  • config_tracker_NvDCF_max_perf.yml: Config file for NvDCF tracker for max perf mode.

  • config_tracker_NvDCF_perf.yml: Config file for NvDCF tracker for perf mode.

  • config_preprocess.txt : Config file for using preprocess in PGIE mode.

  • config_preprocess_sgie.txt : Config file for using preprocess in SGIE mode.

  • source4_1080p_dec_preprocess_infer-resnet_preprocess_sgie_tiled_display_int8.txt : Demonstrates four stream decodes with preprocess plugin in PGIE mode followed by primary inferencing, preprocess plugin in SGIE mode, and three different secondary classifiers. (For dGPU and Jetson AGX Xavier platforms only.)

  • source30_1080p_dec_preprocess_infer-resnet_tiled_display_int8.txt : Demonstrates 30 stream decodes with preprocess plugin in PGIE mode followed by primary inferencing. (For dGPU and Jetson AGX Xavier platforms only.)

  • sources_30.csv: CSV file for 30 sources required in source30_1080p_dec_infer-resnet_tiled_display_int8.yml.

  • sources_4.csv: CSV file for 30 sources required in source4_1080p_dec_infer-resnet_tracker_sgie_tiled_display_int8.yml.

  • source1_usb_dec_infer_resnet_int8.txt: Demonstrates one USB camera as input.

  • source1_csi_dec_infer_resnet_int8.txt: Demonstrates one CSI camera as input. (For Jetson only.)

  • source2_csi_usb_dec_infer_resnet_int8.txt: Demonstrates one CSI camera and one USB camera as inputs. (For Jetson only.)

  • source6_csi_dec_infer_resnet_int8.txt: Demonstrates six CSI cameras as inputs. (For Jetson only.)

  • source2_1080p_dec_infer-resnet_demux_int8.txt: Demonstrates demux mode for two sources.

  • config_mux_source4.txt, config_mux_source30.txt: Sample nvstreammux (new) config files. For more details see Section Mux Config Properties.

• samples/configs/deepstream-app-triton: Configuration files for the reference application for inferencing using Triton Inference Server.

• source30_1080p_dec_infer-resnet_tiled_display_int8.txt (30 Decode + Infer)

• source4_1080p_dec_infer-resnet_tracker_sgie_tiled_display_int8.txt (4 Decode + Infer + SGIE + Tracker)

• source1_primary_classifier.txt (Single source + full frame classification)

Note

Other classification models can be used by changing the ds-triton nvinferserver config file in the [*-gie] group of application config file.

• source1_primary_detector.txt (Single source + object detection using ssd)

• Configuration files for ds-triton nvinferserver element in configs/deepstream-app-triton/

  • config_infer_plan_engine_primary.txt (Primary Object Detector)

  • config_infer_secondary_plan_engine_carcolor.txt (Secondary Car Color Classifier)

  • config_infer_secondary_plan_engine_carmake.txt (Secondary Car Make Classifier)

  • config_infer_secondary_plan_engine_vehicletypes.txt (Secondary Vehicle Type Classifier)

  • config_infer_primary_classifier_densenet_onnx.txt (DenseNet-121 v1.2 classifier)

  • config_infer_primary_classifier_inception_graphdef_postprocessInTriton.txt (TensorFlow Inception v3 classifier - Post processing in Triton)

  • config_infer_primary_classifier_inception_graphdef_postprocessInDS.txt (TensorFlow Inception v3 classifier - Post processing in DeepStream)

  • config_infer_primary_detector_ssd_inception_v2_coco_2018_01_28.txt (TensorFlow SSD Inception V2 Object Detector)

  • config_infer_primary_classifier_mobilenet_v1_graphdef.txt (TensorFlow Mobilenet V1 classifier)

  • config_infer_primary_detector_ssd_mobilenet_v1_coco_2018_01_28.txt (TensorFlow Mobilenet V1 Object Detector)

• samples/configs/deepstream-app-triton-grpc: Configuration files for the reference application for inferencing using Triton Inference Server gRPC

• source30_1080p_dec_infer-resnet_tiled_display_int8.txt (30 Decode + Infer)

• source4_1080p_dec_infer-resnet_tracker_sgie_tiled_display_int8.txt (4 Decode + Infer + SGIE + Tracker)

• Configuration files for ds-triton nvinferserver element in configs/deepstream-app-triton-grpc/

  • config_infer_plan_engine_primary.txt (Primary Object Detector)

  • config_infer_secondary_plan_engine_carcolor.txt (Secondary Car Color Classifier)

  • config_infer_secondary_plan_engine_carmake.txt (Secondary Car Make Classifier)

  • config_infer_secondary_plan_engine_vehicletypes.txt (Secondary Vehicle Type Classifier)

• NVIDIA TAO Toolkit pretrained Models:

  • samples/configs/tao_pretrained_models: Contains README.md to obtain configs and models for TAO toolkit.

• samples: Directory containing sample configuration files, models, and streams to run the sample applications.

• samples/streams: The following streams are provided with the DeepStream SDK:

  Streams	Type of Stream
  sample_1080p_h264.mp4	H264 containerized stream
  sample_1080p_h265.mp4	H265 containerized stream
  sample_720p.h264	H264 elementary stream
  sample_720p.jpg	JPEG image
  sample_720p_mjpeg.mp4	MJPEG containerized stream
  sample_720p.mp4	Containerized stream
  sample_cam5.mp4	H264 containerized stream (360D camera stream)
  sample_cam6.mp4	H264 containerized stream (360D camera stream)
  sample_industrial.jpg	JPEG image
  yoga.jpg	Image for perspective projection in Dewarper
  yoga.mp4	Containerized stream
  sample_qHD.mp4	Used for MaskRCNN
  sample_qHD.h264	H264 elementary stream
  sample_push.mov	H264 containerized stream
  sample_ride_bike.mov	H264 containerized stream
  sample_run.mov	H264 containerized stream
  sample_walk.mov	H264 containerized stream
  fisheye_dist.mp4	Containerized stream
  sonyc_mixed_audio.wav	Audio bitstream
  sample_office.mp4	Containerized stream
  pointcloud	Contains input files for lidar application

• samples/models: The following sample models are provided with the SDK:

DeepStream Reference application

Model	Model Type	No. of Classes	Resolution
Primary Detector	Resnet10	4	640 × 368
Secondary Car Color Classifier	Resnet18	12	224 × 224
Secondary Car Make Classifier	Resnet18	6	224 × 224
Secondary Vehicle Type Classifier	Resnet18	20	224 × 224

Segmentation example

Model	Model Type	No. of Classes	Resolution
Industrial	Resnet18 + UNet	1	512 x 512
Semantic	Resnet18 + UNet	4	512 x 512
Instance	Resnet50 + Maskrcnn	2	1344 x 832

Scripts included along with package

The following scripts are included along with the sample applications package:

• samples/ prepare_classification_test_video.sh: Downloads Imagenet test images and creates a video out of it to test with Classification models like TensorFlow Inception, ONNX DenseNet etc.

• samples/ prepare_ds_triton_model_repo.sh: Prepare the Model repository for Triton Inference Server

  1. Creates engine files for Caffe and UFF based models provided as part of SDK.

  2. Downloads Model files for ONNX DenseNet , SSD Inception V2 Coco, Inception v3.

  For additional information on the above models, refer to:

  • ONNX DenseNet - https://github.com/onnx/models/tree/master/vision/classification/densenet-121

  • SSD Inception V2 Coco - https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/tf1_detection_zoo.md

  • Inception V3 - https://github.com/tensorflow/models/tree/master/research/slim

• uninstall.sh: Used to clean up previous DS installation.