tdioinferenceserver

Architecture 1: Local Annotation Servers & (separate) Inference Server

Architecture 2: Local (combined) Annotation & Inference Server

Architecture 3: Cloud Annotation Server & Inference Server

Download inference server from Github.com

git clone https://www.github.com/trainingdata/tdioinferenceserver
cd tdioinferenceserver

Download models for medical imaging

git submodule update  --init --recursive
cd nvidiaclara
git lfs install
git lfs pull
cd ..

Install & Run

./install.sh
./start.sh

Stop server

./stop.sh

Add ML Model to "ML Models Library"

Create a Labeling Job

As shown in this video https://youtu.be/cjiNQ9DCtLo:

1. Create a labeling job with AI-Assisted annotation.
2. After manually cleaning AI-assisted annotations, launch re-train ML Model

Re-Train ML Model

Re-Run Inferencing