Ashtiani Blink Write2 An Improved Text Entry Method Using Eye Blinks

BlinkWrite2: An Improved Text Entry Method Using Eye Blinks
Behrooz Ashtiani & I. Scott MacKenzie

Dept. of Computer Science and Engineering
York University
Toronto, CANADA
{abehrooz, mack}@cse.yorku.ca

Abstract 1.1 Eye Tracking and Eye Typing
Areas of design improvements for BlinkWrite, an eye blink text Capturing eye biometrics is commonly performed using video
entry system, are examined, implemented, and evaluated. The cameras equipped with infrared or near-infrared light to produce
result, BlinkWrite2, is a text entry system for individuals with corneal reflection. Video-based capturing techniques are more
severe motor impairment. Since the ability to blink is often suitable for use in interactive systems than those based on
preserved, even in severe conditions such as locked-in Electro-OculoGraphy (EOG), scleral contact lens/search coil,
syndrome, BlinkWrite2 allows text entry and correction with Photo-OculoGraphy (POG) or Video-OculoGraphy (VOG)
blinks as the only input modality. Advantages of BlinkWrite2 [Duchowski 2007; Gorodnichy 2004].
over its predecessor include an increase in text entry speed. In a Most video-based eye tracking interfaces, when used for
user evaluation, 12 participants achieved an average text entry computer input, are designed to simulate a standard pointing
rate of 5.3 wpm, representing a 16% increase over BlinkWrite device such as a mouse. A common application is text entry. So
and a 657% increase over the next fastest video-based eye blink called "eye typing" allows the user to enter text by looking at
text entry system reported in the literature. keys on a soft keyboard (a.k.a. an on-screen keyboard or virtual
ACM Classification: K.4.2 [Computer and Society]: Social keyboard) [Majaranta and Räihä 2002]. Keys are typically
Issues – assistive technologies for persons with disabilities; selected by holding the point of gaze (dwelling) on a key for a
H.5.2 [Information Interfaces and Presentation]: User Interfaces predetermined period of time, such as 750 ms.
– input devices and strategies) While existing eye typing systems present a practical method of
Keywords: Blink typing, hands free text-entry, eye typing, communication to some members of the disabled community,
scanning ambiguous keyboard, assistive technologies, locked-in they fail to include the most severely motor impaired. Especially
syndrome, alternative communication, blink input, single input in the case of locked-in syndrome, movement of the eyes is
modality, blink detection limited and often unreliable for use with eye trackers [Laureys et
al. 2005; Leon-Carrion et al. 2002]. As a result, eye typing
1 Introduction systems are not a reliable alternative for persons with these
Humans obtain considerable information through their eyes. For disabilities, where blinking may be the only viable method of
eye typing, the Internet and other digital and information eye-based communication.
technologies continue to enhance our experience as humans,
providing new ways to communicate with friends, colleagues, Fortunately, most eye tracking systems include an adjunct
and loved ones. These technologies have tremendous potential capability of detecting blinks, the rapid closing and opening of
for disabled persons whose conditions hinder their ability to the eyelid. With regard to interface design, the challenge is to
communicate and threaten isolation. Disabilities, such as accommodate text entry using only a single binary input:
amyotrophic lateral sclerosis (ALS), cerebral palsy (CP), or blinking. In this paper, we describe BlinkWrite2, an eye blink
locked-in syndrome (LIS), are severe and often lead to complete text entry system.
loss of control over voluntary muscles, except the eye muscles, 1.2 Eye Blink Text Entry
rendering the individual paralyzed and mute. The Diving Bell and the Butterfly is the memoir of Jean-
The eyes, therefore, become an important input modality to Dominique Bauby, editor in chief of Elle magazine, who entered
improve the quality of life of persons with a severe motor a state of locked-in syndrome in 1995 after suffering a massive
impairment by offering a window of communication to the stroke [Bauby 1998]. Under complete physical paralysis, his
outside world; a goal that is facilitated by enabling text entry only method of communication was through the blink of his left
using the eyes. eye. Entering text for the memoir involved a transcriber reciting
a frequency-ordered alphabet, until he chose a letter by blinking.
The entire book took some 200,000 blinks to compose using this
method. 1 A typical word took two minutes to write (i.e., text
entry speed = 0.5 wpm). Not only is this prohibitively slow, text
composition required two people.
To our knowledge, there are only two examples in the literature
Copyright © 2010 by the Association for Computing Machinery, Inc. of text entry using blinks as the sole input modality; i.e., video-
Permission to make digital or hard copies of part or all of this work for personal or based system without tracking eye movement.
classroom use is granted without fee provided that copies are not made or distributed
for commercial advantage and that copies bear this notice and the full citation on the Su et al. [2008] describe an eye blink-based communication aid
first page. Copyrights for components of this work owned by others than ACM must be for severely disabled individuals. Their proposed system
honored. Abstracting with credit is permitted. To copy otherwise, to republish, to post on
servers, or to redistribute to lists, requires prior specific permission and/or a fee.
1
Request permissions from Permissions Dept, ACM Inc., fax +1 (212) 869-0481 or e-mail http://en.wikipedia.org/wiki/Diving_bell_and_the_butterfly
permissions@acm.org.
ETRA 2010, Austin, TX, March 22 – 24, 2010.
© 2010 ACM 978-1-60558-994-7/10/0003 $10.00

339

consists of task-based menu selection of several daily necessities the selected keys, even if the full word is entered. However, as
such as issuing voiced messages, typing, home appliance long as the word is in the system's dictionary, it is guaranteed to
control, web surfing, etc. Row-column scanning was used to be present in the word-selection region.
increase the rate of item selection, as with other single-switch
input methods [Baljko and Tam 2006; Lesher et al. 1998;
Venkatagiri 1999]. Upon selection of a task, related options or
sub-tasks were displayed on the screen. For the typing task, a
scanning keyboard with three-level selection (a.k.a. block,
group, or quadrant scanning [Bhattacharya et al. 2008; Lin et al.
2007]) and scanning intervals of two seconds was evaluated.
Participants were asked to type a set phrase “ci lab” with their
blinks. A text entry rate of 0.6 words per minute (wpm)2 was
reported.
BlinkWriter [Krolak and Strumillo 2008] replaces the popular
automated scanning method for selection-traversal with a blink-
pattern-based one. Voluntary blinks (duration > 200 ms) are
classified as long blinks and are used for advancing to the next
row and column. A row or column selection occurs if a new
blink is not detected in two seconds. Therefore, typing a letter
involves blinking to reach the desired row, delaying for
selection, blinking to reach the desired column, and a final delay
for letter selection. Letters are arranged based on frequency of
occurrence in Polish. An average entry time of 10.7 seconds per Figure 1. BlinkWrite experimental interface
character, or 0.7 wpm 2 was reported. Importantly, these systems A key aspect of the design of BlinkWrite was choosing the
do not provide an integrated text correction mechanism (i.e., optimal number of letter keys and the letter assignment across
without selection of such a function in a menu outside of the keys. "Optimal" was defined as a design that minimizes the
keyboard itself). average number of scan steps per character (SPC) required to
In the absence of speech, interactive conversation is tolerable enter text in English. In searching for an optimal design, we
only if it achieves a minimum rate of 3 wpm [Darragh and used a frequency-ordered list of the 9,025 most-common words
Witten 1992]. BlinkWrite [MacKenzie and Ashtiani (in press)], in English, as used in previous text entry research [Silfverberg et
the first prototype, met the minimum requirements by reaching al. 2000]. The design in Figure 1 has SPC = 1.713.
novice entry speeds of 4.8 wpm. BlinkWrite2 was developed to
improve that speed. 3 BlinkWrite2
BlinkWrite was developed as an extension of a scanning
In the following section, we provide a brief description of the ambiguous keyboard, inheriting its design principles and
operation of BlinkWrite. This is followed with a description and augmenting it with duration-based blink selection, error
evaluation of the improvements leading to BlinkWrite2. correction, and auditory feedback. BlinkWrite2 is a modified
2 BlinkWrite and improved version of BlinkWrite; these modifications involve
BlinkWrite is a one-key text entry method using a scanning changes to the layout and operation of BlinkWrite. In this
ambiguous keyboard (SAK). It combines the most demanding section, we summarize the operation of BlinkWrite2, noting
requirement of a scanning keyboard (one switch input) with the areas of improvement over the earlier version.
most appealing feature of an ambiguous keyboard (one keypress 3.1 Implementation
per character). BlinkWrite2 captures eye blinks using a TM3 eye tracker from
However, instead of using a physical key, BlinkWrite uses blinks EyeTech Digital Systems (http://www.eyetechds.com/). The
to select virtual keys that are scanned, or highlighted in device is bundled with QuickGlance software that acts as a high
sequence. The interface is shown in Figure 1. level hardware driver and provides general control over eye
tracking, data capture, system setup, and calibration. An
BlinkWrite includes two regions, a letter-selection region and a
additional API, QuickLink, allows further customization of the
word-selection region. The letter selection region includes four
interface and direct access to raw image data. In BlinkWrite, a
virtual keys. The 26 letters of the English alphabet are
background application called TrackerSetup was developed
distributed across three keys. The fourth key is a [space] key.
which modified control variables of QuickGlance to enable
The keys are scanned cyclically with letters selected by blinking
blink clicking and to disable eye movement tracking. However,
when the desired key is highlighted. With each selection, a word
TrackerSetup relied on QuickGlance for blink detection,
list is built from the current key sequence and is presented in the
auditory feedback, and interval durations.
word-selection region. Scanning switches to the word-selection
region when the user selects the [space] key. This is done after QuickGlance provides a large set of features, designed for
selecting all letter keys in a word, or sooner if the desired word general purpose eye tracking; these features, while sufficient for
appears early. Because the letter keys are ambiguous (as with a development of BlinkWrite, impose limitations for BlinkWrite2
mobile phone keypad), there may be multiple words matching implementation. For instance, BlinkWrite2 needs three action
intervals while QuickGlance only supports two action intervals
2
It has been a convention dating back to the 1920s to define "word" in to satisfy its general task of simulating left and right mouse
"words per minute" as five characters, including letters, punctuation, buttons.
etc.[Yamada 1980]

340

In BlinkWrite2, TrackerSetup was replaced by BlinkCapture, an
application we developed to gain direct access to the image data
of the eyes. This approach permitted great flexibility and
allowed code to be written for blink detection, interval
calculations, and auditory feedback. BlinkCapture continually
surveys the image data of the eyes and reports their visibility.
Upon detection of a blink (i.e., both eyes are closed), a timer
starts to record its duration, stopping when at least one eye is
opened.
3.1.1 Blinkwrite2 Prototype Software
In designing BlinkWrite2 to be faster and more efficient, the
model of BlinkWrite was reviewed to find areas of possible
improvement. It became evident that a reduction of SPC
improves the upper limit text entry speed of the system
(discussed in detail in section 3.1.6). This reduction was studied
and applied to both regions of the system: letter-selection and
word-selection.
Within the letter selection region, the [space] key was used as a
method of ending letter-selection and moving to the word-
selection region. However, since the system had to scan over Figure 2. BlinkWrite2's graphical user interface
this key in every cycle, it resulted in unnecessary overhead. Values chosen for lower and upper bounds of other intervals are
Another disadvantage of the [space] key was the extra scans to also based on the results of the pilot study and reflect the
reach it after the letter selection was finished. For instance for a intervals considered most comfortable for issuing commands
word ending in letters “a” to “h”, it took an extra three scans to using blinks. These blinks are named based on their assigned
reach the [space] key and proceed to the word-selection region. action in BlinkWrite2. Blinks of durations between 140 ms and
Further, running all the words in our dictionary through an 540 ms are classified as “selection blinks”. Blink of 540 ms to
algorithm and assuming early word selection opportunities are 1200 ms are “jump blinks”. Blinks longer than 1200 ms are
always taken, 26% of words end on the first key, 28% on second “deletion blinks”.
key, and the remaining 46% on the third key.
The decision to assign the deletion operation to the last interval
With these considerations, we introduced a new command, (blink duration > 1200 ms) is based on results of BlinkWrite
referred to as “jump”, to jump to the word-selection region as prototype which reported an average of 2.2 corrections per
soon as the word needs to be chosen. To access this command, a phrase. Since a jump command is issued at least five times per
new blink interval is implemented in the system. phrase (at the end of typing each word), it is reasonable to assign
Words in the word-selection region were ordered based on it a lower interval to optimize overall speed of the system.
frequency of occurrence. Word selection suffered mainly from
the linear scanning of the list. The word-selection region was 3.1.3 Context Sensitive Commands
redesigned to support column-row scanning of the words to Since the commands are assigned to time intervals, they directly
reduce the scans required to reach the desired word; effectively, affect the performance of the system. As a result, ensuring their
this reduces the overall SPC of the system. However, column- usefulness across the two regions of the interface is important.
row scanning also reduces the overall text entry speed by Further, design decisions must keep the commands simple,
increasing the selections per scan step (SPS), since extra intuitive, and consistent. For instance, the jump operation is used
selections are required to select appropriate rows. The resulting more frequently than delete in the letter-selection region; upon
layout for BlinkWrite2 is shown in Figure 2. discovering that delete is more prevalent in the word-selection
region, it is logical to design the system so that delete is assigned
3.1.2 Blinking Intervals a shorter interval (replace jump’s assigned interval) once
In BlinkWrite2, blinks are classified into four intervals based on scanning moves to the word selection region. This dynamic
duration. To accommodate the new “jump” command, a new assignment of commands to intervals is confusing and
interval is added to the three existing intervals of BlinkWrite and unintuitive to the users who are already cognitively involved in a
all lower and upper bounds are adjusted. timed selection task. With these considerations, we designed the
commands to always remain in their respective intervals but act
Blinks of less than 140 ms are ignored as they are attributed to
according to what is intuitively expected. A summary of actions
involuntary blinks and occasional false positives in blink
and context is provided in Table 1.
detection. This threshold is based on a pilot study to find a
reasonable threshold, separating voluntary and involuntary The select command is used consistently for selection. The new
blinks without sacrificing responsiveness of the system. jump command allows jumping to the word list once letter
BlinkWrite used a 200 ms threshold in accordance with previous selection is finished. While the columns in the word-selection
research [Grauman et al. 2003; Jacob 1990]; however, during the region are scanned, a jump command causes a jump to the first
pilot study this value was found to be too conservative and often column allowing a missed opportunity for column selection to
voluntary blinks, causing frustration. be corrected.

341

Context/Action Select Jump Delete Auditory stimuli are generated at the beginning of intervals of
all commands. For a selection attempt, the feedback is generated
During letter Select a Jump to word Delete a at 140 ms. See Figure 3b. Figure 3c and Figure 3d provide
selection letter list letter/word examples of jump and delete blinks respectively; note how in
During word list Select a Jump to first Delete last addition to the feedback related to their intervals, the feedback
column selection column column letter from previous intervals is also generated. This effect is required
since the action intended by the blink is not known until the
Jump to blink has ended.
During word Select a Delete last
previous
selection word letter
column
Table 1. Summary of command action based on context
Similarly if a column selection was intended but missed, (a)
resulting in the next column being selected, issuing a jump
provides a chance for correction by jumping to the previous
column.
Deleting during letter selection will delete the last letter or the (b)
last word using the method discussed in the next section.
Attempting a delete in the word-selection region means either
the word is mistyped or a jump was issued too early; therefore,
the last typed letter is deleted and scanning returns to the letter
(c)
selection region.
3.1.4 Error Correction
The delete operation remains the same as in BlinkWrite. Both
single-letter delete and word delete are supported. The deletion
(d)
algorithm includes a mechanism to avoid lengthy and frustrating
sequences of single-letter deletes. Briefly, a delete operation
removes either the last keystroke (letter) if input is in the middle Figure 3. Auditory feedback and blink durations.
of a word, or the last word if input is between words. If a single- (a) involuntary blink <140 ms (no feedback) (b) select blink of
letter delete is followed immediately by another delete, the duration 400 ms with auditory feedback at 140 ms (c) jump
effect is to remove the rest of the keystrokes for the current blink of duration 800 ms with auditory feedback at 140 ms and
word. The next delete operation would remove the preceding 540 ms (d) delete blink of duration 1400 ms with auditory
word, and so on. feedback at 140 ms, 540 ms, and 1200 ms
3.1.5 Auditory Feedback The results of BlinkWrite experiments suggested that this poses
In the absence of feedback, users of eye tracking systems are left no problem for the user. In fact, the consistency of having
guessing whether the system captured their intended input distinct auditory signals at the 140 ms, 540 ms, and 1200 ms
correctly. Since input with BlinkWrite2 involves keeping the points helps users accurately learn and perform the available
eyes closed for a period of time, visual feedback can be missed. action commands.
Another problem arises when system commands are performed
based on the duration of blinks. It is difficult for users to adapt 3.1.6 Upper Limit Text Entry Speed.
to and learn the timing by practice. Auditory feedback is used in The timing interval inherent with scanning keyboards places a
BlinkWrite2 in a manner that informs and attunes users of the strict upper limit on entry speed – a limit that is governed by the
exact time an operation is committed. design of the keyboard, rather than a person's proficiency in
using it. For blink-based systems using a scanning ambiguous
BlinkWrite supported only selection and deletion. With the keyboard, as described here, the upper limit is given by:
added complexities of a third action (i.e., jump), it is more
important to utilize sounds that are intuitive (e.g., sound of (1)
recycle bin in the Microsoft Windows XP) and are as distinct as
possible. Since the sounds play at the beginning of each interval, where,
sounds of each preceding interval are played before the sound of
current interval. See Figure 3. It is therefore important for the S = upper limit for text entry speed in wpm
duration of the sounds to be short to avoid overlap. SI = scanning interval in milliseconds
Several sets of sounds were created and subjected to a pilot SPC = number of scan steps per character
study. One set was chosen for BlinkWrite2. The “select” sound SPS = number of selections per scan step
(duration = 30 ms) is softer than a regular mouse click. Jump BT = minimum blink time for selection in milliseconds
and delete sounds (durations = 200 ms) were designed to
BlinkWrite had SPC = 1.713, SPS = 0.627, and BT = 200 ms.
contrast each other (e.g., sounds of insertion and removal of
Assuming SI = 700, BlinkWrite’s upper text entry speed was
USB devices in Microsoft Windows XP).
8.5 wpm. BlinkWrite2’s design modifications and interval
The timing of the auditory feedback with respect to the blink adjustments resulted in SPC = 1.122, SPS = 0.856, and BT = 140
dynamics is illustrated in Figure 3. For completeness, an ms. Therefore, SPC is reduced by 35%, while a negative side
involuntary blink of <140 ms duration is included in the figure effect of 14% increase on SPS is incurred.
(Figure 3a).

342

The upper text entry speed of BlinkWrite2 is: adjusted accordingly, to achieve an optimal lock on their eyes,
evident by a crosshair displayed inside the image of the eyes in
QuickGlance software.
(2)
Participants were allowed to enter a few trial phrases to become
familiar with the selection and correction methods and the
This shows that BlinkWrite2’s upper limit text entry speed is corresponding audio feedback. For the experiment, data
53% higher than that of BlinkWrite. While these values are collection began with the first selection blink for each phrase
promising, a user study is required to determine how the new and ended upon completion of that phrase. The phrases were
system performs in practice. chosen randomly from a standard set of 500 phrases for
evaluating text entry systems [MacKenzie and Soukoreff 2003].
4 Methodology Phrase lengths ranged from 16 characters to 43 characters
(mean = 28.6). A separate scan steps per character analysis was
done of the phrase set. The result was SPC = 1.850, making the
4.1 Participants phrases, on average, more complicated than English in general.
Twelve volunteer participants ranging in age from 22 years to 31 (BlinkWrite2 System has SPC = 1.121) Although, this might
years (mean = 26, SD = 2.7) were recruited. Five were female, seem like a minor point, the phrases used in evaluating systems
seven male. All were daily users of computers, reporting 4 to 10 like BlinkWrite2 play a critical role. For example, it would be
hours of usage per day (mean = 7, SD = 3.4). Three used relatively easy to concoct phrases using words with low SPC
corrective lenses (2 females, 1 male). None had prior experience values, and thereby obtain artificially inflated results. This was
using BlinkWrite or BlinkWrite2. not done. If anything, using a phrase set with a higher SPC than
English will yield results on the conservative side. Participants
4.2 Apparatus were allowed to rest at their discretion between phrases. The
The EyeTech Digital Systems TM3 eye tracker (described QuickGlance software was checked after each break to ensure
above) was connected to a standard Microsoft Windows XP proper blink detection; adjustments were applied if necessary.
notebook with a 2.0 GHz Intel Core2 duo CPU and 4 GB of After completing the experiment, participants were given a
memory. The BlinkWrite2 application was displayed on an questionnaire soliciting demographic information (results cited
external 15" LCD monitor. The eye tracker was raised on a book below) and their impressions of the most suitable scanning
and a notebook case to achieve an optimal angle for eye interval.
detection. The setup is shown in Figure 4.
4.4 Design
The experiment included a single independent variable, scanning
interval, with three levels: 700 ms, 850 ms, and 1000 ms. As we
were interested in determining an optimal scanning interval for
initial use of BlinkWrite2, counter-balancing was used in
assigning the three levels to participants. To offset learning
effects, two participants were assigned to each of the six
possible orders for the three conditions. Values for the scanning
interval were chosen based on results from the earlier BlinkWrite
evaluation [MacKenzie and Ashtiani (in press)]. For each
scanning interval condition, participants entered five phrases of
text, yielding a total of 12 participants × 3 scanning intervals × 5
phrases = 180 phrases of input.
5 Results and Discussion
5.1 Speed
The grand mean for entry speed was 4.71 wpm. This alone is a
promising result, given that entry involves only eye blinks. Of
Figure 4. Experimental apparatus used for BlinkWrite2 the three scanning intervals tested, the nominal setting of 850 ms
evaluation. The EyeTech Digital Systems TM3 is identified by was fastest at 5.30 wpm. The other two scanning intervals were
the block arrow. slightly slower: 4.57 wpm at 1000 ms and 4.27 wpm at 700 ms.
See Figure 5.
4.3 Procedure
Prior to collecting data, the experimenter explained the task and The effect of scanning interval on entry speed was statistically
demonstrated the software. Participants were instructed on the significant (F2,22 = 12.17, p < .0001). Figure 5 also includes the
method of text entry, early word selection, error correction, and upper limit for text entry speed at each scanning interval.
the audio feedback. They were instructed to enter the given The values were calculated using Equation 1 with SPC = 1.122,
phrases as quickly and accurately as possible and make SPS = 0.856, and BT = 140 ms. Although the upper limit
corrections only if an error is detected in the current or previous increases with a decreasing scanning interval, participants were
word. Participants were asked to adjust the height of the chair not able to demonstrate a corresponding increase in text entry
and their distance from the display to positions deemed speed.
comfortable. The angle and distance of the eye tracker were

343

It is noteworthy, however, that with a scanning interval of Errors committed but corrected are reflected in the incidence of
850 ms, participants' text entry speed, at 5.30 wpm, was 48% of delete blinks (> 1200 ms) for letter delete or word delete. The
the upper limit speed of 11.03 wpm and 16% higher than the results are shown in Figure 6. As with entry speed and error
corresponding figure for BlinkWrite. The next fastest eye blink rates, the best performance occurred with a scanning interval of
text entry system, BlinkWriter [Krolak and Strumillo 2008], 850 ms, with 0.78 letter deletes and 0.27 word deletes per phase.
reached the maximum speed of 0.7 wpm. BlinkWrite2 shows a (Recall that the mean phrase length was 28.6 characters,
657% increase in text entry speed when compared to equivalent to 5.7 words.) Delete blinks were three times as
BlinkWriter. prevalent with a scanning interval of 700 ms.
5.3 Optimal Scanning Interval
It seems the three levels chosen for the scanning intervals were
appropriate in the sense that they bracketed user behaviour.
Performance overall was best with a scanning interval of 850
ms: entry speed was fastest, error rate was lowest, and letter
deletes and word deletes were lowest. Participants' performance
suffered at the faster (700 ms) and slower (1000 ms) scanning
intervals. Our observations reaffirm the nominal value of about
850 ms as being preferred for initial use of BlinkWrite2. Of
course with practice, the situation would change. Simpson and
Koester, for example, used an adaptive scanning interval
[Simpson and Koester 1999]. As expertise developed, the
scanning interval was incrementally reduced according to user
performance. Optimal performance was achieved at 525 ms.
With practice, users of BlinkWrite2 could, no doubt, work
effectively with a faster scanning interval and thereby achieve
Figure 5. Text entry speed by scanning interval. Error bars show higher text entry throughput. Of course the scanning interval
±1 SE. with BlinkWrite2 is slightly different than with typical scanning
We consider this a promising result, particularly in view of the keyboards. BlinkWrite2 selections occur through blinks of
fact that each participant entered only five phrases for each duration 140 ms to 540 ms. With each blink, the scanning timer
scanning interval. No doubt, with continued practice entry is restarted, making the full duration of the scanning interval
speeds would increase toward the upper limit. available again for a subsequent selection in the same scan if the
next letter is on the same key. So, faster scanning intervals do
5.2 Accuracy not bring the same sense of "urgency" with BlinkWrite2 as they
The accuracy in participants' responses is revealed in two ways: do with traditional scanning keyboards. Nevertheless, while
(i) the errors in the transcribed text and (ii) the errors committed faster scanning intervals increase text entry throughput with
but corrected. The former were computed using the minimum traditional scanning keyboards, the connection is less direct with
string distance (MSD) method of comparing the presented text BlinkWrite2.
string with the transcribed text string [MacKenzie and Soukoreff
2002]. The error rates are quite low overall, as participants 5.4 Questionnaire
generally chose to correct errors during entry. In fact, 91.12% of The questionnaire solicited participant responses on their level
all phrases had an error rate of 0%. Typically, if any errors were of comfort in using the eye tracker, their eye fatigue, and their
left in the transcribed text, it was because an incorrect word was preference for the scanning interval settings. Responses were
selected early in the phrase or the participant simply didn't look rated on a 7-point Likert scale, where 1 is the least favorable
for or notice the error. Consequently, those transcribed text response and 7 is the most favorable response. The results are
phrases that contained errors, tended to have a large error rate shown in Figure 7.
(e.g., 10% - 20%).

Figure 7. Results of the questionnaire for comfort and
Figure 6. Letter and word deletions per phrase by scanning preference. A response of 7 is most-favourable, and 1 least-
interval. Error bars show ±1 SE. favourable. Error bars = ±1 SE.

344

83% of participants rated the medium scanning interval of 850 LAUREYS, S., PELLAS, F., VAN EECKHOUT, P., GHORBEL, S.,
ms higher than the other two settings. One of the most prevalent SCHNAKERS, C., PERRIN, F., BERRÉ, J., FAYMONVILLE, M.
complaints in BlinkWrite was the fatigue caused by slow E., PANTKE, K. H., DAMAS, F., LAMY, M., MOONEN, G., and
scanning of the candidate word list. Overall, 75% of the users of GOLDMAN, S. 2005. The locked-in syndrome: What is it
BlinkWrite reported a higher than average (average = score of 4) like to be conscious but paralyzed and voiceless?, Progress
eye fatigue during the experiment. By introducing the column- in Brain Research, 150, 495 - 511.
row scanning method for word selection, and more accurate
blink capture by developing BlinkCapture, only 16% of LEON-CARRION, J., VAN EECKHOUT, P., DOMINGUEZ-MORALES,
BlinkWrite2 users reported a higher than average eye fatigue; M. D., and PEREZ-SANTAMARIA, F. J. 2002. The locked-in
67% of users felt comfortable using the eye tracker for syndrome: A syndrome looking for a therapy, Brain Injury,
interaction. 16, 571 - 582.
LESHER, G., MOULTON, B., and HIGGINBOTHAM, J. 1998.
6 Conclusion
Techniques for augmenting scanning communication,
Improvements to the design of BlinkWrite and generic scanning
Augmentative and Alternative Communication, 14, 81 -101.
ambiguous keyboards were proposed. Changes to the layout,
interval and auditory feedback adjustments were implemented as LIN, Y.-L., CHEN, M.-C., WU, Y.-P., YEH, Y.-M., and WANG, H.-
a new prototype called BlinkWrite2. In a user study to determine P., A flexible on-screen keyboard: Dynamically adapting
the practical speed improvements, users achieved a mean text for individuals’ needs, in Universal Access in Human-
entry speed of 5.3 wpm. This represents a 16% increase over Computer Interaction. Applications and Services, Berlin:
BlinkWrite, and a 657% increase over the next fastest eye blink Springer, 2007, 371 - 379.
text entry system reported in the literature. It is expected that
BlinkWrite2 holds promise as a viable text entry method for MACKENZIE, I. S. and ASHTIANI, B. BlinkWrite: Efficient text
users with severe motor disabilities who are otherwise precluded entry using eye blinks, Universal Access in the Information
from using computing technology for communication. Society (in press).
MACKENZIE, I. S. and SOUKOREFF, R. W. 2002. A character-level
7 References
error analysis technique for evaluating text entry methods.
BALJKO, M. and TAM, A. 2006. Indirect text entry using one or
In Proceedings of the Second Nordic Conference on
two keys. In Proceedings of the ACM SIGACCESS
Human-computer Interaction – NordiCHI 2002, New York:
Conference on Assistive Technologies - ASSETS 2006, New
ACM, 243 - 246.
York: ACM, 18 - 25.
MACKENZIE, I. S. and SOUKOREFF, R. W. 2003. Phrase sets for
BAUBY, J. D., The diving bell and the butterfly. New York:
evaluating text entry techniques, in Extended Abstracts of
Vintage International, 1998.
the ACM Conference on Human Factors in Computing
BHATTACHARYA, S., SAMANTA, D., and BASU, A. 2008. User Systems - CHI 2003, New York: ACM, 754 - 755.
errors on scanning keyboards: Empirical study, model and
MAJARANTA, P. and RÄIHÄ, K.-J. 2002. Twenty years of eye
design principles, Interacting with Computers, 20, 406 -
typing: Systems and design issues. In Proceedings of the
418.
ACM Symposium on Eye Tracking Research and
DARRAGH, J. J. and WITTEN, I. H., The reactive keyboard. Applications - ETRA 2002, New York: ACM, 15 - 22.
Cambridge: Cambridge University Press, 1992.
SILFVERBERG, M., MACKENZIE, I. S., and KORHONEN, P. 2000.
DUCHOWSKI, A., Eye tracking methodology theory and practice Predicting text entry speed on mobile phones. In
2nd edition, London: Springer, 2007. Proceedings of the SIGCHI Conference on Human Factors
in Computing Systems – CHI 2000, New York: ACM, 9 -
GORODNICHY, D. O. 2004. Towards automatic retrieval of blink- 16.
based lexicon for persons suffered from brain-stem injury
using video cameras. In Proceedings of the 2004 SIMPSON, R. C. and KOESTER, H. H. 1999. Adaptive one-switch
Conference on Computer Vision and Pattern Recognition row-column scanning, IEEE Transactions on
Workshop - CVPRW 2004, New York: IEEE, 68 - 68. Rehabilitation Engineering, 7, 464 - 473.
GRAUMAN, K., BETKE, M., LOMBARDI, J., GIPS, J., and BRADSKI, SU, M., YEH, C., LIN, S., WANG, P., and HOU, S. 2008. An
G. R. 2003. Communication via eye blinks and eyebrow implementation of an eye-blink-based communication aid
raises: Video-based human-computer interfaces, Universal for people with severe disabilities. In International
Access in the Information Society – UAIS 2003, 2, 359 - Conference on Audio, Language and Image Processing,
373. 2008 - ICALIP 2008 , New York: IEEE, 351 - 356.
JACOB, R. J. K. 1990. What you look at is what you get: Eye VENKATAGIRI, H. S. 1999. Efficient keyboard layouts for
movement-based interaction techniques. In Proceedings of sequential access in augmentative and alternative
the SIGCHI Conference on Human Factors in Computing communication, Augmentative and Alternative
Systems - CHI 1990 , New York: ACM, 11 - 18 Communication, 15, 126 - 134.
KROLAK, A., and STRUMILLO, P. 2008. Vision-based eye blink YAMADA, H. 1980. A historical study of typewriters and typing
monitoring system for human-computer interfacing. In methods: From the position of planning Japanese parallels,
Conference on Human System Interactions – HIS 2008, Journal of Information Processing, 2, 175 -202.
New York: IEEE, 994 - 998.

345

Ashtiani Blink Write2 An Improved Text Entry Method Using Eye Blinks

More Related Content

Viewers also liked (20)

Similar to Ashtiani Blink Write2 An Improved Text Entry Method Using Eye Blinks (20)

More from Kalle (20)

Recently uploaded (20)

Ashtiani Blink Write2 An Improved Text Entry Method Using Eye Blinks