DeepStream Getting Started
-
Highlights:
- Integration with Triton Inference Server (previously TensorRT Inference Server) enables developers to deploy a model natively in TensorFlow, TensorFlow-TensorRT, PyTorch, or ONNX in the DeepStream pipeline
- Smart recording on edge
- Python development support with sample apps
- Build and deploy apps natively through RHEL
- Secure communication between edge and cloud using SASL/Plain based authentication and TLS authentication
- IoT Capabilities
- DeepStream app control from edge or cloud with bi-directional IoT messaging
- Dynamic AI model update on the go to reduce app downtime
- Interoperability with models from Transfer Learning Toolkit 2.0
- Jetson Xavier NX support
Jetson | T4 (x86) | |
---|---|---|
Operating System | Ubuntu 18.04 |
Ubuntu 18.04 RHEL 8 |
Dependencies |
CUDA: 10.2 cuDNN: 8.0.0 TensorRT: 7.1.0 JetPack: 4.4 |
CUDA: 10.2 cuDNN: 7.6.5+ TensorRT: 7.0.0 Driver: R440+ |
Getting Started Resources
Downloads
DeepStream 4.0 applications are fully compatible with DeepStream 5.0. Please read the migration guide for more information.
Python Sample Apps & Bindings
Python bindings is now integrated in the DeepStream SDK
Visit the DeepStream Python Apps Github page for documentation and sample apps.
Check out the DeepStream SDK technical FAQ for questions commonly asked.
FAQ
- Check out the frequently asked questions on DeepStream SDK in the technical FAQ
Documentation & Forums
- Release Notes
- Getting Started Guide
- Migrate from prior DeepStream versions to DeepStream 5.0: see documentation
- DeepStream Python API
- Plug-ins manual
- Post your questions or feedback in the DeepStream SDK developer forums
Reference Implementations
- DeepStream Python Apps: repository contains Python bindings and sample applications
- DeepStream C/C++ Apps: repository contains C/C++ bindings and sample applications
- Using DeepStream SDK for Redaction: reference implementation shows how to redact faces and license plates in video streams
- Add/delete source at RuntimeRuntime source addition/deletion application to show the capability of Deepstream SDK
- Anomaly detectionThis project contains anomaly detection application and auxiliary plug-ins to show the capability of Deepstream SDK.
- Back-to-back detectors This project contains a Back to Back detector application to show the capability of Deepstream SDK.
- DeepStream apps with TLT models : This repository provides a DeepStream sample application to run six Transfer Learning Toolkit models (DetectNet_v2 / Faster-RCNN / YoloV3 / SSD / DSSD / RetinaNet).
- Using custom YOLO models in DeepStream : The objectDetector_Yolo sample app provides a working example of the open source YOLO models such as YOLOv2, YOLOv3, tiny YOLOv2, and tiny YOLOv3
- 360d smart parking application with DeepStream: An end to end smart parking application implementation using DeepStream SDK
Blogs & Tutorials
- Creating a Human Pose Estimation Application with NVIDIA DeepStream
- Deploying Models from TensorFlow model zoo Using NVIDIA DeepStream and NVIDIA Triton Inference Server
- Implementing a Real-time, AI-Based, Face Mask Detector Application for COVID-19
- Building Intelligent Video Analytics Apps Using NVIDIA DeepStream 5.0
- Video tutorial DeepStream best practices for unlocking greater performance
- Building a Real-time Redaction App using NVIDIA DeepStream Part 1 | Part 2
- Bringing Cloud-Native Agility to Edge AI Devices with the NVIDIA Jetson Xavier NX Developer Kit
- Improving INT8 accuracy using Quantization Aware Training and NVIDIA Transfer Learning Toolkit
- Developer Tutorial: Training Instance Segmentation Models using MaskRCNN on NVIDIA Transfer Learning Toolkit
- Training with Purpose-built Pre-trained Models Using the NVIDIA Transfer Learning Toolkit
- How to integrate NVIDIA DeepStream on Jetson Modules with AWS IoT Core and AWS IoT Greengrass
- DeepStream edge-to-cloud with Azure IoT
AI Training with Transfer Learning Toolkit
DeepStream Turnkey Integration with Cloud Services
Perception & Analytics
Beginner Friendly Free Self-Paced DLI Online Courses
- Learn how to build end-to-end intelligent video analytics pipelines using DeepStream and Jetson Nano >> Enroll now
- Learn how to get started with AI using Jetson Nano >> Enroll now
Webinars
- GTC 2020: Deepstream with Open-Horizon (Presented by IBM)
- GTC 2020: Implementing Real-time Vision AI Apps Using NVIDIA DeepStream SDK
- Create intelligent places using NVIDIA pre-trained vision models and DeepStream SDK
- Build with DeepStream, deploy and manage with AWS IoT services
- Intelligent Video Analytics using DeepStream 5.0
- Build and deploy AI-powered intelligent streaming analytic apps with DeepStream SDK
- GTC 2020: NVIDIA Tools to Train, Build, and Deploy Intelligent Vision Applications at the Edge
- GTC 2020: How to Build a Multi-Camera Media Server for AI Processing on Jetson
- GTC 2020: Google Cloud AutoML Video and Edge Deployment
- GTC 2020: Visual Anomaly Detection using NVIDIA Deepstream IoT Workshop
- GTC 2020: Productionizing GPU Accelerated IoT Workloads at the Edge
- DeepStream edge-to-cloud with Azure IoT
- DeepStream SDK- Accelerating Real-Time AI Based Video And Image Analytics
- Real-time Object Detection for Disaster Response
Additional Resources
NVIDIA’s platforms and application frameworks enable developers to build a wide array of AI applications. Consider potential algorithmic bias when choosing or creating the models being deployed. Work with the model’s developer to ensure that it meets the requirements for the relevant industry and use case; that the necessary instruction and documentation are provided to understand error rates, confidence intervals, and results; and that the model is being used under the conditions and in the manner intended.