This tool allows collecting gaze data necessary for personal calibration or training of eye-tracking models. It was developed as part of my master's thesis on eye tracking with a monocular webcam. The framework for the full gaze tracking pipeline is also available.
The output is a folder with a CSV file containing the target that the person is looking at in pixels and the file name of the associated webcam image. For good calibration results, it is recommended to take at least 9 calibration images, the more, the better.
pip install -r requirements.txt
python calibrate_camera.py
, see Camera Calibration by OpenCV.python main.py --base_path=./data/p00
--monitor_mm=750,420 --monitor_pixels=1920,1080
, and adjust the TargetOrientation
values in utils.py
.E
is pointing at when the letter color changes from blue to orange. Please press the arrow key several times because sometimes OpenCV doesn't register the click the first time.q
key when the data collection is complete.python visualization.py --base_path=./data/p00
.