YOLOv3 gRPC Extension for AVA Edge, Inc.

YOLOv3 gRPC Extension for AVA Edge, Inc.

YOLOv3 model inferencing based on ONNXRuntime (CPU, GPU)

This module is developed from 「IoT Edge module extensions to be used with Azure Video Analyzer」 repository to be able to run on IoT Edge.

While running the IoT Edge module, you can use in conjunction with Azure Video Analyzer. It can be used as an inferencing server for gRPC extension node in Azure Video Analyzer Edge. The performance of the YOLOv3 model inferencing server depends on CPU or GPU power.

This module can be used with the Azure cloud native application WeDX Flow to simplify module management.

Minimum hardware requirements:

  • YOLOv3 gRPC-CPU Extension : Linux x64 and arm64 OS, 2GB of RAM, 1.5GB of storage
  • YOLOv3 gRPC-GPU Extension : Linux x64(CUDA >= 11.4) and arm64(JetPack >= 4.4) OS, 3GB of RAM, 3GB of storage


  • Inferencing server for AVA gRPC extension node
  • Video stream of inference results - http://{IoT Edge Device IP}/stream
  • gRPC Server Address - tcp://{Module Name}:8151
  • Support shared memory

Direct methods

  • Not available

Environment variables

  • Not available

Desired properties

  • Stream Video of inference results (default:true)
    • "ViewVideoStream": true / false
  • Debug Output File results (default:false)
    • "DebugOutputFile": true / false

AVA(Azure Video Analyzer) Integration

  • Pipeline Topology sample (GrpcExtension)
    • "url": "tcp://{Module Name}:8151"
    • "mode": "pad"
    • "width": "416"
    • "height": "416"
    • "@type": "#Microsoft.VideoAnalyzer.ImageFormatJpeg"

Inference results

  • http://{IoT Edge Device IP}/stream
    • Viewing the output video with inferencing overlays in a browser

RTSP Publishing for Android App