Train defect detectors from paired factory images.
What it does
Trains, evaluates, and exports models that detect manufacturing defects by comparing image pairs.
When to use it
Use when training, evaluating, exporting, or running inference with a TAO Optical Inspection model on quality-control data.
How to use it
Give it inspection images and CSV metadata; it runs model training and saves the results.
What you provide
- Your files
Uses
TT
Good to know
- a credential is hard-coded: encryption_key
Latest review
What you need · 6
Requires Docker for the container-based workflow.
Requires NVIDIA Container Toolkit alongside Docker.
Requires at least one GPU with 8GB or more VRAM; one GPU is recommended.
Uses Python 3 to run the required dataset preflight before training, evaluation, or inference.
Access · 0
No additional actions listed.