prosdk-addons-dotnet - Calibration Validation

What is it

Addons for the Tobii Pro SDK.

The Tobii Pro SDK can be found here.
The Tobii Pro SDK .Net documentation can be found here.

As of this writing the addon contains functionality for calibration validation. Do not hesitate to contribute to this project and create issues if you find something that might be wrong or could be improved.

Get the addon

Either download the addon and unzip it, or clone the repository:

git clone https://github.com/tobiipro/prosdk-addons-dotnet.git

Building

• Open the Tobii.Research.Addons.sln in Visual Studio. The solution is located in the source folder.
• Add the Tobii Pro SDK to the project via NuGet. For this example we use the 32 bit version Tobii.Research.x86.
• Build the addons assembly. In this example we build the debug version. It ends up in \source\Tobii.Research.Addons\bin\Debug\Tobii.Research.Addons.dll.

Usage example

The following is an example test application that makes use of the calibration validation implementation in the addons library. Note: this example does not display any points to perform a proper calibration validation, it only shows how to use the addons library.

• In Visual Studio, create a Console App (.Net Framework) for Visual C#.
• Add the Tobii Pro SDK to the project via NuGet. For this example we use the 32 bit version Tobii.Research.x86.
• Add a reference to the Tobii.Research.Addons.dll, for example from the Building section above.
• Add the necessary using statements to be able to use the Tobii Pro SDK and the addons library:

using Tobii.Research;
using Tobii.Research.Addons;

• Get a reference to your eye tracker. Here we assume that we have only one connected and get the first one found. Print the address to the console to verify that we got hold of it:

var eyeTracker = EyeTrackingOperations.FindAllEyeTrackers().FirstOrDefault();
Console.WriteLine("Found eye tracker {0}", eyeTracker.Address);

• Create a ScreenBasedCalibrationValidation object and provide the eye tracker reference as an argument:

var calibrationValidation = new ScreenBasedCalibrationValidation(eyeTracker);

• The constructor also accepts arguments for the number of samples to collect for each point, default 30, and the timeout in milliseconds, default 1000.
• Create the points to be used for validation:

var points = new NormalizedPoint2D[] {
    new NormalizedPoint2D(0.1f, 0.1f),
    new NormalizedPoint2D(0.1f, 0.9f),
    new NormalizedPoint2D(0.5f, 0.5f),
    new NormalizedPoint2D(0.9f, 0.1f),
    new NormalizedPoint2D(0.9f, 0.9f)
};

• Enter the validation mode. This makes the calibration validation object start listening to gaze data from the eye tracker:

calibrationValidation.EnterValidationMode();

• Loop through the points and provide them to the validation object. For each point, wait for the data collection to finish. In a real application here is where the points would be displayed on the screen:

foreach (var point in points)
{
    Console.WriteLine("Collecting for point {0}, {1}", point.X, point.Y);

    calibrationValidation.StartCollectingData(point);
    while (calibrationValidation.State == ScreenBasedCalibrationValidation.ValidationState.CollectingData)
    {
        System.Threading.Thread.Sleep(25);
    }
}

• Compute the result:

var result = calibrationValidation.Compute();

• The result object contains the calculated average accuracy and precision and the list of points where each point contains the accuracy and precision for that point plus the data collected for that point. The results are in degrees. In this example we simply print the results:

Console.WriteLine(calibrationValidation);

• When satisfied with the results, call LeaveValidationMode. This clears all collected data and makes the calibration object stop listening to the eye tracker.

calibrationValidation.LeaveValidationMode();

• All implementation details for the ScreenBasedCalibrationValidation class can be viewed in the source code.
• Here is the entire example at once:

using System;
using System.Linq;
using Tobii.Research;
using Tobii.Research.Addons;

namespace TestAddons
{
    internal class Program
    {
        private static void Main(string[] args)
        {
            var eyeTracker = EyeTrackingOperations.FindAllEyeTrackers().FirstOrDefault();
            Console.WriteLine("Found eye tracker {0}", eyeTracker.Address);

            var calibrationValidation = new ScreenBasedCalibrationValidation(eyeTracker);

            var points = new NormalizedPoint2D[] {
                new NormalizedPoint2D(0.1f, 0.1f),
                new NormalizedPoint2D(0.1f, 0.9f),
                new NormalizedPoint2D(0.5f, 0.5f),
                new NormalizedPoint2D(0.9f, 0.1f),
                new NormalizedPoint2D(0.9f, 0.9f)
            };

            calibrationValidation.EnterValidationMode();

            foreach (var point in points)
            {
                Console.WriteLine("Collecting for point {0}, {1}", point.X, point.Y);

                calibrationValidation.StartCollectingData(point);
                while (calibrationValidation.State == ScreenBasedCalibrationValidation.ValidationState.CollectingData)
                {
                    System.Threading.Thread.Sleep(25);
                }
            }

            var result = calibrationValidation.Compute();
            Console.WriteLine(calibrationValidation);
            calibrationValidation.LeaveValidationMode();
        }
    }
}