MCPcopy Create free account
hub / github.com/NVIDIA-AI-IOT/deepstream_reference_apps

github.com/NVIDIA-AI-IOT/deepstream_reference_apps @DS_8.0

Chat with this repo
repository ↗ · DeepWiki ↗ · release DS_8.0 ↗ · + Follow
407 symbols 810 edges 61 files 101 documented · 25% updated 2mo ago★ 1,339

Browse by type

Functions 326 Types & classes 81
What it actually does AI analysis from the code graph — generated when you open this
loading…
README

Reference Apps using DeepStream 8.0

This repository contains the reference applications for video analytics tasks using TensorRT and DeepSTream SDK 8.0.

Getting Started

We currently provide three different reference applications:

Preferably clone this project in /opt/nvidia/deepstream/deepstream/sources/apps/sample_apps/

To clone the project in the above directory, sudo permission may be required.

For further details, please see each project's README.

Anomaly Detection : README

The project contains auxiliary dsdirection plugin to show the capability of DeepstreamSDK in anomaly detection. sample anomaly output

Runtime Source Addition Deletion: README

The project demonstrates addition and deletion of video sources in a live Deepstream pipeline.

MaskTracker: README

This sample app demonstrates DeepStream MaskTracker for multi-object tracking and segmentation using SAM2. sample MaskTracker output

Single-View 3D Tracking: README

The sample app demonstrates single-view 3D tracking with DeepStream multi-object tracking to reconstruct 3D human model in world coordinates under occlusion. sample 3D tracking output

Multi-View 3D Tracking: README

The samples demonstrate multi-view 3D tracking in DeepStream, a distributed, real-time framework designed for large-scale, calibrated camera networks.

sample mulit-view 3D tracking output sample BEV output from multi-view 3D tracking

Parallel Multiple Models Inferencing: README

The project demonstrate how to implement multiple models inferencing in parallel with DeepStream APIs.

Bodypose 3D Model Inferencing: README

The sample demonstrate how to customize the multiple input layers model preprocessing and the customization of the bodypose 3D model postprocessing. Bodypose 3D sample output

Video Buffers sharing between pipelines through IPC: README

This sample demonstrates how to share video buffers over IPC and how to change output video buffers.

Legacy DeepStream Reference Samples

Some old samples are not supported any more for different reasons. The legacy samples are moved to the legacy folder /legacy_apps.

Core symbols most depended-on inside this repo

Shape

Function 262
Class 75
Method 64
Enum 6

Languages

C++62%
C20%
Python18%

Modules by API surface

deepstream-bodypose-3d/sources/deepstream_pose_estimation_app.cpp42 symbols
legacy_apps/deepstream-occupancy-analytics/deepstream_test5_app_main.c24 symbols
deepstream_parallel_inference_app/tritonclient/sample/apps/deepstream-parallel-infer/deepstream_parallel_infer_app.cpp24 symbols
deepstream-tracker-3d-multi-view/utils/deepstream_auto_configurator.py20 symbols
deepstream-ipc-test-sr/deepstream_ipc_test_app.c19 symbols
legacy_apps/deepstream-retail-analytics/nvdsmeta_schema.h17 symbols
legacy_apps/deepstream-retail-analytics/retail_iva.c15 symbols
deepstream-ipc-test-sr/video_template_impl/vt_impl.cpp14 symbols
deepstream-bodypose-3d/sources/deepstream-sdk/eventmsg_payload.cpp14 symbols
legacy_apps/deepstream-retail-analytics/nvmsgconv/deepstream_schema/deepstream_schema.cpp13 symbols
legacy_apps/deepstream-occupancy-analytics/includes/nvdsmeta_schema.h13 symbols
deepstream-tracker-3d-multi-view/utils/kafka_bev_visualizer.py13 symbols

For agents

$ claude mcp add deepstream_reference_apps \
  -- python -m otcore.mcp_server <graph>

⬇ download graph artifact

Ask about this repo answers extend the page