Skip to main content

Overview

Why Use Ioptiq​

The Ioptiq application transforms vision models from isolated inspection tools into a transparent, controllable, and continuously improving quality system. It enables you to:

  • Gain full visibility into all inspection results across your production lines
  • Eliminate black-box behavior by making every model prediction traceable and reviewable
  • Accelerate the feedback loop between operations and the vision team
  • Continuously improve model accuracy through structured feedback collections
  • Maintain control over multiple vision setups within one centralized environment The result: faster model optimization, improved detection performance, reduced manual validation effort, and higher confidence in automated quality control. iOPTIQ Inspections overview with a grid of inspected macaron tray images and filters on the left

Inspections​

Upon opening the application, users are presented with an overview of inspections for the active model. This overview allows you to:

  • View all inspection results generated by the model
  • Filter inspections by a selected date range (start and end date)
  • Filter inspections based on the predictions found on each image
  • Option to view the results in a list or grid with thumbnails to have a quick overview of what the model is generating.

This provides operational insight into model performance and inspection throughput.

Detailed Prediction Analysis​

For in-depth analysis, users can open any individual inspection image. Within the detailed view, you can:

  • See all predictions made by the model on that specific image
  • Analyze detected features or anomalies
  • Review prediction confidence and results per case This enables detailed case-by-case validation and performance assessment. Inspection detail view of a macaron tray with labelled bounding boxes and a Predictions list with confidence scores

Collections​

If the model produces an incorrect or uncertain prediction, the image can be added to a Collection. A Collection is a structured group of images that can be reviewed by the vision team to:

  • Process operational feedback efficiently
  • Analyze recurring false positives or false negatives
  • Improve and retrain the model in a targeted manner

Wrong Prediction collection in iOPTIQ with five inspection images and a Download collection button

Adding Images to a Collection​

Images can be added in two ways:

Via the Inspection Overview

  • Select one or multiple images using the checkboxes
  • Add them to an existing collection or create a new collection

Via the Detail View

  • Open a specific image
  • Add it directly to a collection

This structured feedback mechanism ensures a short and efficient improvement cycle.

Consulting collections​

All created collections are accessible through the Collections Overview. This provides:

  • A structured overview of bundled feedback cases
  • Clear traceability of reported model issues
  • A central reference point for optimization discussions

Sources​

In sources it is possible to enable / disable and change all aspects of your configuration in the UI itself.

  • Rename image fields and metadata
  • Reorder images & metadata
  • Enable / Disable predictions, images & metadata
info

Adding or deleting items is currently only possible via the json configuration. Info about this can be found here: Configuration

Projects​

If multiple vision models or production setups are active on-site, separate projects can be configured. A project represents one complete vision setup on the production floor. Users can:

  • Switch between projects via the project selector at the top of the application
  • Maintain clear separation between different production lines or model configurations

Go to Configuration to see how to setup your very own project.