Home Forums Software Development Specifications regarding analytical use

Viewing 13 posts - 1 through 13 (of 13 total)
• Author
Posts
• #8789
Rasa Bhattarai
Participant

Hey all,

I am about to purchase tobii for analytical use. Before I do, I have few queries. Can anyone explain me about few things I have queries about ? I do not know if this is the platform to ask these questions. I asked in other community forums but have not got any replies yet.

1) What is saccade amplitude ? (Is it the degree traveled by the eye gaze points every second) ?
2) If saccade amplitude is the angular distance traveled (or degrees traveled per second), doesn’t it mean same as saccade velocity (Since velocity is distance traveled per second ?)
3) What other eye tracking features can be evaluated for eye experiment ?

#8791
Grant [Tobii]
Keymaster

Hi @rasa, thanks for your query

1) What is saccade amplitude ? (Is it the degree traveled by the eye gaze points every second) ?

No, the amplitude of a saccade is defined as the angular distance the eye travels between two fixations. (It’s a measure of distance, not speed)

2) If saccade amplitude is the angular distance traveled (or degrees traveled per second), doesn’t it mean same as saccade velocity (Since velocity is distance traveled per second ?)

the Saccade Velocity is the Saccade Amplitude divided by the time interval.

3) What other eye tracking features can be evaluated for eye experiment ?

Many other features indeed! We have created an online article that explain in more depth the types of eve movements and what they might mean for analysis. You can read the article @ https://www.tobiipro.com/learn-and-support/learn/eye-tracking-essentials/types-of-eye-movements/

If you would care to elaborate on your project intentions, we should be able to provide more specific information.

Should you have any further questions, please let us know how we can be of help.

#8793
Rasa Bhattarai
Participant

Hey,

Yes, I have read that saccade amplitude is the angular distance traveled by the eye. But in the papers that I read it mentions the saccade amplitude in terms of degrees. Is it necessary to measure the amplitude in degrees ?

Thanks

#8794
Grant [Tobii]
Keymaster

Hi @rasa, as the distance is *angular* the only other option that I would say other than degrees would be radians, so not much difference really. Please let me know of any other queries you might have.

#8795
Rasa Bhattarai
Participant

Hi @grant,

Thank you for allowing me to ask other queries.

Would you explain how to calculate ‘Saccade amplitude‘ for such gaze points ? Because i want to find out ‘Saccade velocity‘ as shown in this Article and as you mentioned Saccade velocity = Saccade amplitude / time interval.
Thank you again.

#8798
Rasa Bhattarai
Participant

Hello,

From the Article, it specifies that it calculated saccade velocity and acceleration using two point differentiator. From the position(degree) vs time figure in this article, I tried to calculate saccade velocity using the method specified in this Stackexchange platform. (Since the platform question specifies ‘Use a set of data points from a graph to find a derivative’).

Based on the method, I tried to manually calculate(thinking there might be variation of approx few degrees) using the formula specified, the value I obtained is : (10-0)deg / (0.3-0.1)sec = 50 deg/second. But for the transition from 0.2 to 0.3 in velocity vs time graph, it shows reaching 400 deg/second. I am confused regarding this. How is it possible to reach such huge value ? Is there any problem in my calculation ?

Sorry for such detailed question in the tobii platform. But it is very clear to me from your explanation.

Thank you

#8800
Grant [Tobii]
Keymaster

Hi @rasa, whilst it is a nice idea, unfortunately the derivate of the raw gaze locations against their respective time points is not a reliable way to calculate saccadic amplitude as this metric relies on the distance travelled between gaze *fixations*.

To determine therefore Saccadic Amplitude within your program you would need to integrate a fixation filter yourself from the literature in the code and thereafter determine the distance travelled between the two fixation points over a time period.

A number of third parties have created Matlab fixation filters which perhaps you can perhaps translate for your purposes in Unity:
https://se.mathworks.com/matlabcentral/fileexchange/56236-djangraw-gazevistoolbox
& one in R:
https://github.com/davebraze/FDBeye/wiki/Researcher-Contributed-Eye-Tracking-Tools

In any event, you need to be aware that any analysis of data undertaken using the EyeX and associated SDK’s (Unity, Core SDK, etc) in expressly prohibited without the purchase of a special licence.

#8812
Rasa Bhattarai
Participant

Hello @grant,

Thank you for the clarification as always 🙂 May be integrating the third party application would help to determine the fixation points.

Just one clarification: After we obtain the fixation points, Is the saccade amplitude the distance between the two fixation points ? (As you mentioned in the above post: “The amplitude of a saccade is defined as the angular distance the eye travels during the movement. (It’s a measure of distance, not speed)”
Similar to the above mentioned answer: Can saccade velocity be derived from dividing the thus obtained saccade amplitude by time or is it needed to use 2 point differentiation ?

Thank you

#8814
Grant [Tobii]
Keymaster

Hi @rasa, glad to hear you are finding the information useful!

Can saccade velocity be derived from dividing the thus obtained saccade amplitude by time or is it needed to use 2 point differentiation?

Yes exactly, the saccadic velocity is simply the saccadic amplitude divided by the time interval between the two fixation points.

Let me know any other points.

#8821
Rasa Bhattarai
Participant

Hello @grant,

Thank you for allowing me to ask queries regarding eye tracking. I came across one third party open source software which is used for analyzing post eye tracking points. I ran it’s sample program (along with the data points that it provided) and it was regarding fixation_detection. I want to learn how do these fixation points give out saccade amplitude and velocity (Similar to these graphs? The values that is obtained in the output of the program is as such:

Fixations:
Total Number of Fixations(t1,t2,minDur): 22
Total Number of Fixations(t1,3s,minDur): 22

t1,t2,minDur:
ID-Xcenter-Ycenter-Nt1-Nt2-StartTime-EndTime-Duration
1 689.5723 363.6341 70 70 1000.3780 3350.5640 2350.1860
2 1037.1145 367.1904 27 27 3451.1380 4350.4660 899.3280
3 684.5436 716.0460 28 28 4450.2700 5381.6600 931.3900
4 326.7230 361.5158 27 27 5481.6370 6348.1720 866.5350
5 680.4622 -2.4757 30 29 6448.7920 7414.7770 965.9850
6 680.0611 357.9217 28 28 7514.9370 8414.8190 899.8820
7 1026.4849 360.3656 29 28 8548.6230 9448.3750 899.7520
8 674.2867 714.6757 28 27 9548.4600 10415.1300 866.6700
9 326.8549 355.5990 28 28 10514.6500 11414.0100 899.3600
10 627.2755 26.5071 7 7 11481.5300 11680.9700 199.4400
11 665.6554 13.0811 24 24 11714.7600 12481.1100 766.3500
12 681.1126 362.2682 30 30 12573.5100 13514.4300 940.9200
13 667.8339 13.6558 30 28 13580.7600 14547.6200 966.8600
14 684.5231 719.7679 30 30 14613.8500 15580.9300 967.0800
15 319.6370 359.0985 29 29 15647.5700 16580.6700 933.1000
16 1036.1421 360.5191 26 26 16747.2700 17580.6900 833.4200
17 672.2159 354.2256 29 29 17647.3300 18580.8400 933.5100
18 1034.1954 348.7120 30 30 18647.3000 19613.7400 966.4400
19 689.1827 713.7524 29 28 19746.9800 20647.1400 900.1600
20 327.0446 345.6037 28 28 20747.3200 21647.1000 899.7800
21 682.1914 -8.3242 29 29 21747.4100 22680.6700 933.2600
22 689.3789 364.1872 9 9 22747.5300 23014.5400 267.0100

These x and y center shows the fixation centers in pixels/screen coordinates. I tried using the method you said to calculate saccade velocity. But I do not understand how is it that researchers use degree/second to calculate it ? I mean the value that is given here is shown in pixels. So how do I convert it to degrees/second to calculate saccade velocity ?

I am sorry if it is too much to ask. I did not get reply from other sources.

Thank you

#8823
Grant [Tobii]
Keymaster

Hi @rasa, good to hear that you managed to find a tool that will help you determine fixations based on the raw gaze data. Would be useful if you could share the source please?

But I do not understand how is it that researchers use degree/second to calculate it ? I mean the value that is given here is shown in pixels. So how do I convert it to degrees/second to calculate saccade velocity ?

So, this is similar to what was discussed in a previous post @ https://developer.tobii.com/community/forums/topic/accuracy-in-tobii-eye-x/
where one would first calculate the linear distance between two set of pixels co-ordinates using pythagoras theorem.. what is missing from the data above is the distance between the screen and the user. Once you have both the linear distance travelled between two fixations and the distance between screen and user you would determine the angle travelled as = tan-1 (linear distance between fixations / distance between screen and user) and so therefore the sacaddic velocity would be this distance divided by the time elapsed between each fixation.

Hopefully this is clear, let me know.

#8826
Rasa Bhattarai
Participant

Hello @grant,

Using the method you stated for the calculation, I tried using these steps [Taking distance between user and screen = 60cm] :
1. For points (2 and 3) -> (x1, y1) = (1037.115, 367.1904) at time(t = 3.35) and (x2, y2) = (684.5436, 716.046) at time (t = 4.45), I used pythagoras theorem as you stated and found the value as 495.990 px which changed to 13.12 cm during conversion. This gave the value of theta = arctan(13.12/60) = 12.35 degrees [Assuming a standard distance of 60cm]
2. Now, from my understanding 12.35 degrees is the saccade amplitude (angular distance eye traveled during this fixation movement). So is the saccade velocity = 12.35degrees/(4.45-3.45)seconds = 12.35 degrees/seconds ?

Is my calculation wrong somewhere ? Usually according to many papers saccade velocity begins from 100 degree/seconds. Would you help me point out my mistake and clear my confusion?

Thank you sir again.

#8831
Grant [Tobii]
Keymaster

Hi @rasa, yes the technique is correct although I would question that the data you are working with is truly a fixation and not simple raw gaze data.

You are quite correct that the values of saccade velocity are typically higher, so I would check the source of how you determine fixation…can you pass me the link and I will take a look for you.

I found this tool which may be of used to you? https://www.tobiipro.com/about/partners-resellers/app-market/fixation-detector/

Viewing 13 posts - 1 through 13 (of 13 total)
• You must be logged in to reply to this topic.