This toolbox provides a realtime visualization of eye tracking data to be used for designing passive BCI system.
Project description
IMPORTANT
In case, this toolbox doesn't works as intended or you find some issues, create an issue in the tracker (https://github.com/VinayIN/pyETA-toolbox/issues) and I will try to resolve it. (Please attach errors & log)
pyETA-toolbox (Eye Tracker Analyzer)
- This provides a real-time eyetracking metric (especially fixation & Saccades).
- This tool also provides for validation of the eye-tracker performance on a user.
- The tool provides
22 channels, following table describes the EEG channel information in order:
| No. | Channel Name | Type | Unit |
|---|---|---|---|
| Left Eye | |||
| 1 | left_gaze_x |
gaze | normalized |
| 2 | left_gaze_y |
gaze | normalized |
| 3 | left_pupil_diameter |
pupil | mm |
| 4 | left_fixated |
fixation | boolean |
| 5 | left_velocity |
velocity | px |
| 6 | left_fixation_timestamp |
timestamp | s |
| 7 | left_fixation_elapsed |
duration | s |
| 8 | left_filtered_gaze_x |
filtered_gaze | normalized |
| 9 | left_filtered_gaze_y |
filtered_gaze | normalized |
| Right Eye | |||
| 10 | right_gaze_x |
gaze | normalized |
| 11 | right_gaze_y |
gaze | normalized |
| 12 | right_pupil_diameter |
pupil | mm |
| 13 | right_fixated |
fixation | boolean |
| 14 | right_velocity |
velocity | px |
| 15 | right_fixation_timestamp |
timestamp | s |
| 16 | right_fixation_elapsed |
duration | s |
| 17 | right_filtered_gaze_x |
filtered_gaze | normalized |
| 18 | right_filtered_gaze_y |
filtered_gaze | normalized |
| Screen Data | |||
| 19 | screen_width |
screen | px |
| 20 | screen_height |
screen | px |
| 21 | timestamp |
timestamp | s |
| 22 | local_clock |
timestamp | s |
Alternative Installations
Download the pyeta-toolbox-<version>-py3-none-any.whl file from here: https://github.com/VinayIN/pyETA-toolbox/releases and install using pip
# move the .whl file to your workspace
# once moved verify using ls bash command, if found
# replace .whl file with the one just downloaded. it should have this format with a different <version>
pip install pyeta-toolbox-<version>-py3-none-any.whl
or
CAUTION: This below method installs the development version and might have bugs
pip install git+https://github.com/VinayIN/pyETA-toolbox.git
Usage
There are 2 ways this tool can be used (CLI - track/window/validate & GUI - browser/application), both are shown below:
prerequisite: pyETA-toolbox needs to be installed
1. CLI
Command:
pyETA track
Parameters:
--push_stream:Pushes the data to an LSL stream.--data_rate:Specifies the rate of the data stream.--use_mock:Uses a mock service for the eye tracker.--fixation:Adds fixation duration to the data stream.--velocity:Specifies the velocity threshold for fixation.--accept_screen_nans:Avoids correcting for NaNs.--save_data:Saves the data to a file.--verbose:Displays debug statements.--duration:Specifies the duration for which to track the data.--screen_index:for multi-screen execution, use the index to capture the data from a specific screen with the resolution
An example with tracker running with fixation and a mock service that runs for a duration of 10sec and stops (if --use_mock is not provided, it searches for the eye tracker)
pyETA track --fixation --use_mock --duration 10
Command:
pyETA window
Parameters:
--use_mock:Use mockup tracker--screen_index:Screen index to display the validation window--verbose:Enable verbose logging
Example with parameters to run the validation window
pyETA window --use_mock --screen_index=0
Command:
pyETA validate
Parameters:
--csv:Name of the output CSV file. (example: myfile.csv)
Example with parameters to fetch the metrics
pyETA validate --csv=myfile.csv
2. GUI
Command:
pyETA application
or execute the pyETA.exe(To be implemented)
