DL Streamer Pipeline Server

• Overview and Architecture
  • Architecture Overview
    • High-Level Architecture Diagram
  • Supporting Resources
• System Requirements
  • Supported Platforms
  • Minimum Requirements
  • Software Requirements
  • Validation
  • Supporting Resources
• Get Started
  • Prerequisites
  • Quick try out
    • Pull the image and start container
    • Run default sample
  • Legal Information
  • Advanced Setup Options
  • Troubleshooting
  • Known Issues
  • Contact Us
  • Supporting Resources
• Troubleshooting
  • Using REST API in Image Ingestor mode has low first inference latency
  • Axis RTSP camera freezes or pipeline stops
  • Deploying with Intel GPU K8S Extension
  • Deploying without Intel GPU K8S Extension
  • Using RTSP/WebRTC streaming, S3_write or MQTT fails with GPU elements in pipeline
  • RTSP streaming fails if you are using udfloader
  • Resolving Time Sync Issues in Prometheus
  • WebRTC Stream on web browser
  • Error Logs

How to

• How to change Deep Learning Streamer pipeline
  • Steps
• How to use GPU for decode and inference
  • Pre-requisites
    • Ensure you have a GPU
    • Provide GPU access to the container
    • Hardware specific encoder/decoders
    • GPU specific element properties
  • Tutorial on how to use GPU specific pipelines
• How to use CPU for decode and inference
  • Tutorial on how to use CPU specific pipelines
• How to autostart pipelines
  • Steps
    • Method 1 - Specifying all the information in the pipeline itself -
    • Method 2 - Add REST payload under "payload" section of the pipeline config.
• How to launch configurable pipelines
• How to perform WebRTC frame streaming
• How to publish metadata and frame over MQTT
• How to publish frames to S3
  • Steps
• How to publish metadata to InfluxDB
  • Steps
• How to publish metadata over ROS2
  • Prerequisite
  • Publish
  • Subscribe (Example)
• How to launch and manage pipeline (via script)
  • Steps
• How to use RTSP camera as a source
  • RTSP Cameras
  • Using RTSP as a source
• How to run User Defined Function (UDF) pipelines
  • Prerequisite
  • Steps
• How to Deploy with Helm
  • Prerequisites
  • Quick try out
    • Pull the helm chart (Optional)
    • Configure and update the environment variables
    • Install the helm chart
    • Run default sample
  • Troubleshooting
  • Summary
  • Learn More
  • Legal Information
• How to use image file as source over REST payload
  • Image file as source
    • Asynchronous vs Synchronous behavior
    • Async mode
    • Sync mode
• How to download and run YOLO models
  • Steps
    • Download
      • Step 1: Create virtual environment
      • Step 2: Activate virtual environment
      • Step 3: Upgrade pip to latest version
      • Step 4: Download and install packages
      • Step 5: Download Yolo11 model
      • Step 6: Deactivate virtual environment
    • Run YOLO model
• How to build from source
  • Steps
    • Prerequisites
    • Build DL Streamer Pipeline Server image and start container
    • Run default sample
  • Learn More
• How to add system timestamps to metadata
  • Steps

References

• API Reference
  • Markdown
• Environment Variables
  • Mandatory
    • Enable and specify port for REST server
    • RTSP related config
    • Username and ID
    • proxy settings
  • Optional
    • MQTT related configs
    • S3 related settings (Configure only if S3 integration is enabled)
    • OPCUA related configuration (Configure only if OPCUA is enabled)
    • Open Telemetry related config (Configure only if open telemetry is enabled)
    • WebRTC related config (Configure only if WebRTC is enabled)
    • InfluxDB related config (Configure only if InfluxDB is enabled)
    • Miscellaneous env variables

• Release Notes