Human in the Loop

Let WizyVision improve the performance of your ML thanks to human feedback.

After deployment, it’s important to have humans review the reliability and accuracy of custom object recognition.

WizyVision has 2 levels of human interaction with Machine learning:

Level 1 – End-users in the loop

Through accessing the WizyVision homepage, users review whether an object is in an image or not. It’s simple to give feedback: one click per image.

Level 2 – Experts in the loop

Where automatic detections have occurred, your experts have the ability to make corrections thanks to WizyVision’s advanced image annotation interface:

  • Images can be reviewed and marked as incorrect
  • Any image can be reviewed.

Once the review of a set of images is complete, experts can re-train the model to improve its performance if required.

WizyVision. Best of humans for the best of ML.
Image annotation

Allow your experts to annotate images easily in a dedicated, no-code interface.

Human in the loop

Set up an advanced human feedback process to improve the performance of your custom object recognition.

Humans in the feedback loop

Helping machine learning improve its performance is just a matter of clicks.