TAUCHI – Tampere Unit for Computer-Human Interaction Visualizing gaze path for analysis Oleg Špakov MUMIN workshop 2002, Tampere.

TAUCHI – Tampere Unit for Computer-Human Interaction

Visualizing gaze path for analysis

Oleg Špakov

MUMIN workshop 2002, Tampere


Content

• Introduction (6 pages)

• Visualizing method (8 pages)

• Interaction method (8 pages)

• Conclusions (2 pages)

1


1.1 Goals

• Develop the gaze-capturing and gaze-visualizing system for further development as a template

• Write components (drivers) for system-device communication and database creation

• Visualize gaze-path as a set of interactive objects

• Create an environment to run experiment tasks

• Write some common tasks

I. Introduction

2


1.2 Gaze path

• It is a collection of instantaneous dots on 2D surface

• All of them constrain 3 different objects: fixations, saccades and smooth pursuit

• Could be interrupted by blink

• Usually measured in degrees according to basic spot (0,0), which is a normal vertex to face flat

• Huge amount of data needs to be analyzed3


1.3 Fixations

Fixation is

• A part of gaze path, where an eye didn’t make big movements and there are no jumps

• Obtained as an average of X and Y coordinates of constituents dots

• Also it is a part of gaze path, where an eye made very small movements (directed or not). The size of the movement is conventional.

4


1.4 Saccades

• Saccade is a fast eye movement

• Saccades could be purposive (usually long) and involuntary (short and frequent)

• Saccades of the second type usually are automatically filtered by gaze-capturing devices

• Details of captured (by eyes) visual image are unavailable during saccade

5


1.5 Smooth pursuit

• Smooth pursuit is a slow eye movement, when the gaze velocity is higher than for fixation, and lower than for saccade.

• Velocity always waves, and smooth pursuit often interrupted by correlating saccades

• Relatively hard to recognize

6


1.6 Blinks

• Always generate 2 saccades – down at the beginning and up at the end

• Usually interrupts fixation

• There no data from gaze-capturing device during blink – the position of eye is unavailable

7


II. Visualization methods

• Always contains timestamp and gaze projection position (X,Y) to the screen

• Can contain pupil size of eye(s)

• Can contain head-tracking data

• Can contain data for both eyes

2.1 Gaze path data

8


2.2 Main visualization method

• All data presents in fixations and saccades, which are visible interactive objects

• Blinks are the property of saccades

• All fixations and saccades distributed on the 2D surface

• Every fixation and saccade object visualize its properties

9



Fixation is the circle with properties:• Radius• Border width• Border color• Background color• Central spot size

Saccade is the line with properties:• Width• Color

10


Visualized data for both fixations and saccades

• Duration

• Time

• Average pupil size

Visualized data for saccades only:

• Blinks number


11



Duration > Radius Duration > Border ColorDuration > Back-Color

12

Duration > Line width


2.3 Data over background

• Allow to overview the observed image parts

• Allow to makes links between images and gaze

13


2.4 Data as a clipping region

• Shows only the observed part of image

• Calculates rates of observation time

14


2.5 Samples’ visualization

• Many samples for every object requires a lot of resources

• Only samples of 1 currently active fixation or/and saccade are visualized

15


3.1 Navigation• Visible

• Interactive

• Adaptive

• Directed

• Through fixations, saccades and its samples

III. Interaction methods

16


3.2 Selection

• Different criterions

• Interactive

• Visible

• Different selection modes

17


18

3.2 Selection


3.2 Selection

19


3.2 Selection

20


21

3.2 Selection


3.3 Data obtaining

Application allows

• to get object- and raw-data of any fixation, saccade and blink

• to get object- and raw-data of any selection of fixations

• to save any object- and raw-data in a file

• to load and save any selection range

22


3.4 Additional features

• Image loading to background

• Zooming and panning of visible layout

• Full-screen mode

• User-task implementation interface

23


4.1 Conclusions• Different kinds of gaze path visualization

were developed

• Different tools for navigation and selection were developed

• Visualization parameters are adjustable by user

• Researchers can use it as a tool to analyze gaze path

• Also can be used as a template for further development and user tasks creation

• Some user tasks were created24


4.2 Future steps

• Improve time-position-time navigation

• Add extra features to events objects

• Create 3D navigation

• Movie on the background

25


Thank you for attention

Everyone is inviting to take part in

development, consulting and testing

[email protected]

http://www.cs.uta.fi/~oleg

TAUCHI – Tampere Unit for Computer-Human Interaction Visualizing gaze path for analysis Oleg Špakov MUMIN workshop 2002, Tampere.

Documents