SIMULATION

specifications:

The tracking volume of faceLAB 4 is flexible and can be adjusted to accommodate your specific experimental requirements. The standard field-of-view allows very precise gaze tracking, or long distance tracking. The field-of-view can be expanded to allow even larger head movements. Modes are configured or changed through a simple software ‘wizard’.

Using multiple faceLAB systems and faceLAB-LINK, tracking volume can be expanded to a size limited only by the number of faceLAB systems that are linked.

Smallest Face

• Automatic tracking initialisation when face is only 20 percent of total image width;


• 6 DOF head tracking and recovery when face is only 10 percent of total image width.

Largest Face

• Continues to track when only half of the face is visible.

Head Rotations

• Tracking and recovery up to +/- 90° around neck axis (turn head from shoulder to shoulder);
  
• Tracking and recovery up to +/- 45° around nod axis (look up / look down);
  
• Tracking up to +/- 120° and recovery up to+/- 30° around tilt axis (lean left / right).

Gaze Rotations

• Eye rotations of +/- 45°.

Recovery Time

• Tracking recovery times approximately 100ms for both head and eye measurements when subject returns to the field-of-view.

Obscuration

• Tracking when up to 50% of face is obscured;
  
• Recovery requires 80% of face to be visible.

Precision Gaze Configuration†
This configuration tracks head position and produces the highest accuracy data on eye gaze direction, without restricting head movement. Precision mode is recommended for applications such as on-screen usability or stimulus studies, simulator and psychology experiments.

• Camera field of view out to 30°;
  
• Freedom of head motion (single system): horizontal tracking range up to 0.35m (13.8”), vertical range up to 0.23m (9”), distance range up to 0.6m (23.6”);
  
• Head tracking distance range from 0.4 to 1.4m;
  
• Typical static accuracy of gaze direction measurement 0.5-1° rotational error, depending on selected field-of-view;
  
• Pupil diameter measurement (independent left/right eyes);
  
• Eye vergence distance (meters).

faceLAB "Classic" Configuration†
In this configuration, both head-pose and gaze tracking are possible. Gaze calibration is not required, but can be performed to remove any systematic bias.

The specifications are similar to previous versions of faceLAB, with the exception that tracking recovery and robustness to large rotations is greatly improved. Classic mode is recommended for applications such as on-road intelligent vehicles studies, aerospace installations and fatigue studies.

• Camera field of view out to 30°;
  
• Freedom of head motion (single system): horizontal tracking range up to 0.35m (13.8”), vertical range up to 0.23m (9”), distance range up to 0.6m (23.6”);
  
• Head tracking distance range from 0.4 to 1.4m;
  
• Typical static accuracy of head measurement within +/- 1mm of translational error and +/- 1° of rotational error;
  
• Typical static accuracy of gaze direction measurement within +/- 5° rotational error.

Ultra-Wide Field-Of-View Configuration
In this configuration, head-position and rotation can be automatically tracked over large volumes without calibration.

• Camera field of view out to 45°;
  
• Head tracking distance range from 0.5 to 1.4m.

Stimulus

• Customisable number of stimulus points for on-screen calibration;
  
• Customisable type of stimulus (both audio and visual) for on-screen calibration.

Eye Analysis Measurements

• Gaze direction measurement;
  
• PERCLOS fatigue assessment (day and night operation);
  
• Blink detection, including blink artefact correction for gaze direction, pupil diameter and vergence distance;
  
• Blink rate (Hz), and mean period (secs);
  
• Saccade detection.

Analysis and Data Transmission

• Compatible with Eye Response Technologies GazeTracker™ analysis tool suite for stimulus presentation, data collection & analysis;
  
• Simultaneous real-time and filtered tracking data available as user-definable disk log, network transmission or serial transfer (using SDK);
  
• Real-time output to screen provides comprehensive feedback to the operator;
  
• 3D OpenGL World Model Editor/Viewer to assess interaction with environment;
  
• Time-stamped ASCII output also available for use with common applications such as Microsoft Excel™, Matlab™, SPSS™ etc;
  
• Experiment-time synchronisation across multiple systems available via NTP;
  
• WorldView data visualisation software for data playback, recording and modelling.

Support

• Comprehensive illustrated colour manual;
  
• Online help, web-based user group support services, phone and email support;
  
• Portable software development kit (SDK) available for ease of data integration into custom applications/devices or simulators;
  
• On-site customer training is also provided.

Vehicle and Simulator Versions
Customisable packs are available for easy incorporation into test-cars, aircraft/vehicle simulators and immersive display areas. Real-time and analysis integration available from simulation vendors such as Oktal, Drivesafety and 5DT.

† Example test results available at www.seeingmachines.com.
* Laptop hardware and specification may vary from pictured.