WWW.THESIS.XLIBX.INFO
FREE ELECTRONIC LIBRARY - Thesis, documentation, books
 
<< HOME
CONTACTS



Pages:     | 1 |   ...   | 5 | 6 || 8 | 9 |   ...   | 10 |

«A dissertation submitted to the Department Of Computer Science, Faculty of Science at the University Of Cape Town in partial fulfilment of the ...»

-- [ Page 7 ] --

The two questions in the pilot study covering the details of the video were extended to five questions, to include motion blurring, the speed of the video in addition to the facial and hand detail visibility.

A fifth question was added to determine if a low score on intelligibility is purely because of the quality of the video or if unfamiliarity of a Sign Language phrase were impacting on the scoring of the video clips.

Question 4.1 Possible answer The movement was clear. The movement was blurry.

From the comments in the pilot study, blurred video was often a problem in the clips. This question was added in response to these comments, it is focussed on the movement of the hands and arms being blurred, something that is expected to happen at lower frame rates.

The answer to this question was captured as a numeric value, with 5 being given to the movement was clear and a value of 1 to the movement was blurry.

Question 4.2 Possible answer I could clearly see all the details of I had difficulty seeing the details the face. of the face.

This question was present in the pilot study, but has been simplified in this study to a binary answer.

Sign Language uses two main parts of the body for communications, the face as well as the hands of the speaker. Question 4.2 and 4.3 focuses on these two areas and attempt to evaluate the impact lowering the frame rate and resolution has on the comprehension of these areas separately.

Question 4.2 focused on the face of the speaker.

The answer to this question was captured as a numeric value, with 5 being given to I could clearly see all the details of the face and a value of 1 to I had difficulty seeing the details of the face.

Question 4.3 Possible answer I could clearly see the hands. I had difficulty seeing the hands.

This question was present in the pilot study, but has been simplified in this study to a binary answer.

Sign Language uses two main parts of the body for communications, the face as well as the hands of the speaker. Question 4.2 and 4.3 focuses on these two areas and attempt to evaluate the impact lowering the frame rate and resolution has on the comprehension of these areas separately.

Question 4.3 focused on the hands of the speaker.

The answer to this question was captured as a numeric value, with 5 being given to I could clearly see the hands and a value of 1 to I had difficulty seeing the hands.

Question 4.4 Possible answer The video was the right speed. The video was too slow/too fast.

This question, as is the case with question 4.1, was added in response to the comments participants made during the pilot study. The two main complaints were blurred motion and the speed of the video clip being wrong.

This and blurring of motion is a function of the frame rate of the video clip. The lower the video clip’s frame rate the lower the rating should be for questions 4.1 and 4.4.

The answer to this question was captured as a numeric value, with 5 being given to the video was the right speed and a value of 1 to the video as too slow/too fast.

Question 4.5 Possible answer I knew all the signs. Some signs were unknown to me.

The pilot brought another question to mind. If a participant finds one of the Sign Language phrases unfamiliar, what impact will that have on their evaluation of the intelligibility of the video clip? This question was added in response.

The answer to this question was captured as a numeric value, with 5 being given to I knew all the signs and a value of 1 to some signs were unknown to me.

Question 5 How many times did you view this clip?

It was decided in this experiment to do away with the single view of video clip constraint and rather provide the participant the opportunity to review the clip as needed, but record the number of views on the questionnaire.

The single view constraint was removed to more closely resemble the conversational use of a Sign Language video clip where the listener could ask the signer to resign the previous phrase.

Question 6 Any other comments on this video?

Question 6 provided the participant the opportunity to give any general comments on the just viewed and evaluated video clip.

As with question 1, the answer to this question was captured, but no numeric value was assigned to the answer.

4.3 Observations With only six clips used during this experiment, instead of the 12 as in the pilot, there was no problem with oddly ordered clips (A1, A10, A11, A12, A2, A3 … A8, A9) and all video clips were selected and viewed without any problems. Again all participants were clearly familiar and comfortable using the cell phones.

Again, as with the first pilot study, the participants were willing to write down their responses in English.

Where the assistance of the SASL interpreter was needed though, was helping with the correct spelling of words, and in a few cases the English word for a sign. The words in these cases were simply finger spelled out for the participant.

–  –  –

Table 4-2 contains the mean participant rating for each video clip, as well as the ANOVA significance value for each of the seven questions as well as for the average participant rating over all the questions. As can be seen in the table all of the questions returned a significance level of greater than 0.05 (p 0.05) and, therefore, there is no statistically significant difference in the mean participant rating for each of the video clips. No combination of frame rate and video resolution, either high or low, was preferred significantly more or less than any other combination of frame rate and resolution.





Figure 4-1 to Figure 4-7 show the average participant rating for each of the questions answered by the participants in the questionnaire, with Figure 4-8 showing the overall average participant rating across all questions.

Figure 4-1: Qualitative results for Question 2. Figure 4-2: Qualitative results for Question 3.

The qualitative results for the question “How sure are you The qualitative results for the question “How easy or how of your answers to Question 1 above?” for each of the difficult was it to understand what was said in this video?” for each of the three frame rates and two three frame rates and two resolutions. With a significance level of 0.732 (p =.732) there was no statistically resolutions. With a significance level of 0.840 (p =.840) significant difference in the average participant rating for there was no statistically significant difference in the each of the video clips. average participant rating for each of the video clips.

Figure 4-3: Qualitative results for Question 4.1. Figure 4-4: Qualitative results for Question 4.2.

The qualitative results for the question “The movement The qualitative results for the question “I could clearly see was clear/blurry.” for each of the three frame rates and all the details of the face/I had difficulty seeing the details two resolutions. With a significance level of 0.172 (p = of the face.” for each of the three frame rates and two.172) there was no statistically significant difference in the resolutions. With a significance level of 0.825 (p =.825) average participant rating for each of the video clips. there was no statistically significant difference in the average participant rating for each of the video clips.

Figure 4-5: Qualitative results for Question 4.3. Figure 4-6: Qualitative results for Question 4.4.

The qualitative results for the question “I could clearly see The qualitative results for the question “The video was the the hands/I had difficulty seeing the hands.” for each of right speed/The video was too slow/too fast.” for each of the three frame rates and two resolutions. With a the three frame rates and two resolutions. With a significance level of 0.491 (p =.491) there was no significance level of 0.208 (p =.208) there was no statistically significant difference in the average statistically significant difference in the average participant rating for each of the video clips. participant rating for each of the video clips.

–  –  –

Figure 4-8: Overall mean participant response across all questions.

The overall mean participant response and standard deviation across all questions for each of the three frame rates and two resolutions. The y-axis is the average participant response. Each group on the x-axis is a particular video resolution, with each colour representing a particular video frame rate. With a significance level of 0.540 (p =.540) there was no statistically significant difference in the average participant rating for each of the video clips.

Again, as with the pilot study, there were no statistically significant results. The video clip at a resolution 176 x 144, at 15 frames per second, were the only clip from the six clips that all six participants agreed that the movement was clear, could see clearly all the details in the face, could clearly see the hands, and the video was at the right speed.

The impact of compression was eliminated from this experiment by compressing the video clips with a bit rate specification far above the required bit rate for the given resolutions and frame rates.

Looking at the final bit rate values listed in Table 4-1, all final bit rates are all well below the specified bit rate of 5000 kbits/sec, with the highest being only 2663.28 kbit/sec.

Where the comments made by the participants in the first pilot study hinted at problems with unknown signs, the follow-up study’s comments confirmed that the actual signs used in the test

video clips are impacting on the intelligibility of the videos:

“some sign language confuse” “boy – different sign language” “sign language bit confuse” “problem with sign play” “problem with sign boy” “No, it was difficult about sign language” Only one clip (176 x 144 at 30 frames per second) was confirmed to contain only known signs by the participants. Whether the rest of the clips truly contained signs unknown to the viewer or it was simply a case of not being able to identify the sign because the sign was not clear enough in the video, is not clear from the results.

This experiment added a question to the questionnaire to test the possibility of unknown signs impacting on the subjective evaluation of the intelligibility of the video clips. The response to Question 4.5 (shown in Figure 4-7), as well as the comments from participants points to the different SASL dialects leading to a single sign making perfect sense to some participants, while being completely unknown to other participants.

Again, as in the first pilot study, no statistically significant results were recorded. Two possible reasons may be the small sample group, and the fact that the same participant evaluated multiple clips of different resolutions and frame rates. The same participant viewing and evaluating multiple clips at different specifications could impact the participant evaluation of subsequent video clips.

5 Intelligibility Study (Experiment 3) Based on the results and gathered feedback of the two pilot studies the final intelligibility study was conducted (see Appendix C).

5.1 Aim The final intelligibility study aimed to incorporate all the lessons learnt through the two preceding pilot studies to answer the main research question: What is the lowest video resolution and frame rate that would provide intelligible SASL video on a cell phone?

Through the experience and findings of the pilot studies the following changes were made to the

experimental setup in the final intelligibility study:

 Both pilot studies gave no statistically significant results. In an effort to increase the chances of the final experiment giving a statistically significant result the number of participants was greatly increased. In addition, to remove the possibility that a participant’s response to a specific video clip could be impacted by a previously viewed video clip, each participant only watched and gave feedback on a single video clip. With multiple participants evaluating the same video clip.

 The binary questions of the follow-up study could not give a clear enough picture of the participants’ opinions and were replaced with a consistent set of five-level Likert items all using the typical Likert scale, this not only simplified the explanation of the questionnaire to the participants, but also the analysis of the answers. It was now possible to directly compare answers and form a true overview value.

 The question statements in this study were based on the statements used in the binary questions of the second pilot study. Each set of binary statements used in the follow-up study became two Likert scale statements in this study.

 The two pilot studies both gave statistically insignificant results, with no clear preference or rejection of any of the resolution-frame rate combinations. In an effort to attain statistical significance the number of resolution-frame rate combinations being compared was reduced.

With the number of resolutions being evaluated already at only two, the number of different frame rates was reduced. With the objective of finding the lowest usable frame rate and resolution supporting intelligible SASL video communications the 30 frames per second frame rate was dropped focusing on the lower end of the frame rate scale. In a further effort to attain statistical significance 15 frames per second was replaced by 20 frames per second to have a more distinct difference in frame rate between the video clips.

 To minimise the possibility of the actual signed contents of the clips having an impact on the evaluation of the video clip, only signed phrases with no known dialectic differences were used.

 Lastly following the results of the original pilot study and the subsequent changes to the preparation of the video clips, the compression and cropping of the video clips were kept the same as for the second pilot study.

5.2 Procedure



Pages:     | 1 |   ...   | 5 | 6 || 8 | 9 |   ...   | 10 |


Similar works:

«wetter grimma wetter grimma Wetter Grimma, Sachsen, Deutschland. 16-Tages Das Wetter in Grimma. Wettertrend für die nächsten 16 Tage inkl. Temperatur, Wetterzustand und Regenwahrscheinlichkeit. 14-Tage-Wetter Grimma WetterOnline Wetter Grimma. Das 14-Tage-Wetter für die Region Grimma mit den weiteren Wetteraussichten zu Sonne, Niederschlag, Wind, und Temperaturen. Wird es wärmer Wetter Grimma | wetter.com Wie wird das Wetter heute in Grimma Temperatur-, Windund Regenvorhersage, sowie...»

«LOCOMOTIVE SAFETY STANDARDS The Locomotive Inspection Act makes it unlawful for any carrier to use or permit to be used on its line any locomotive unless the entire locomotive and its appurtenances (1) are in proper condition and safe to operate in the service for which they are put, without unnecessary peril to life or limb and (2) have been inspected and tested as required by the regulations. When a locomotive has one or more conditions not in compliance, it may be moved only as a light...»

«Echt Interim B.V. Koningin Anna Paulownastraat 17 8019 XN Zwolle [ T ] 038 853 54 15 [ M ] 06 30 00 38 28 [ E ] info@echtinterim.nl [ I ] www.echtinterim.nl Triodos Bank 39.04.61.814 KvK 08167805 BTW-nr. NL821906835B01 Bouwer Tikkeltje eigenzinnig Recht door zee, zeer betrokken en uiterst effectief Resultaatgericht met veel humor Energiek Curriculum vitae 9 pagina‟s Personalia Naam : Van Echten Voornamen : Jolanda Adres : Koningin Anna Paulownastraat 17 Postcode : 8019 XN Woonplaats : Zwolle...»

«Chapter 2. Editing And Compiling Now that the main concepts of programming have been explained, it’s time to actually do some programming. In order for you to “edit” and “compile” a program, you’ll need a program to type, an editor, and a compiler program. At first, what you will type will be provided for you – you’ll type exactly what this book tells you to type. But as we move forward, you’ll have more and more of an opportunity to write parts yourself, and ultimately write...»

«TRAINING MODULE ON PERSONALITY DEVELOPMENT Sponsored by Department of Personnel & Training Government of India & U.N.D.P Prepared by Dr.Chandra Prasad Sreedhar & Mr.Oommen Mathew Institute of Management in Government Thiruvananthapuram FOREWORD The influence of human personality upon the functional efficiency of an organization and its personnel has been widely recognized. The personality can also be modified to a certain extent. Accordingly, recent years have seen a variety of efforts by...»

«The Responsible Investment Practices of the World’s Largest Government Sponsored Investment Funds By: Hugues Létourneau A thesis submitted to the Faculty of Graduate and Post-Doctoral Affairs in partial fulfillment of the requirements for the degree of M.A. In International Affairs Carleton University, Ottawa, Canada © 2013 Hugues Létourneau Abstract: In this thesis, I evaluate the role of agent related (e.g.: assets under management, type of fund) and structure related (e.g.: strength of...»

«WP/13/9 Banks’ Foreign Credit Exposures and Borrowers’ Rollover Risks: Measurement, Evolution and Determinants Eugenio Cerutti © 2013 International Monetary Fund WP/13/9 IMF Working Paper Research Department Banks’ Foreign Credit Exposures and Borrowers’ Rollover Risks: Measurement, Evolution and Determinants 1 Prepared by Eugenio Cerutti Authorized for distribution by Stijn Claessens January 2013 This Working Paper should not be reported as representing the views of the IMF. The views...»

«Should Gold-Exporters Peg Their Currencies to Gold? A study for the World Gold Council Jeffrey Frankel Harpel Professor, Harvard University First draft: January 27; revised May 27, 2002 Summary The debate over the best choices of monetary standards and exchange rate regimes for developing countries is as wide open as it has ever been. On the one hand, the big selling points of floating exchange rates – monetary independence and accommodation of terms of trade shocks – have not lived up to...»

«I Introduction Page 1. Introduction 5 1. What is bankruptcy? 7 2. When should people file bankruptcy? 9 3. How is this book organized? 11 II All About Assets 1. Main residence 14 2. Other real estate 16 3. Cars and trucks 17 4. Bank accounts 19 5. Stocks, bonds, mutual funds and other investments 20 6. Retirement accounts 21 7. Life insurance policies 23 8. Other personal property (furniture, jewelry, firearms, etc.) 9. Small businesses 29 10. Personal injury, workers compensation, and social...»

«Regulatory Circular RG14-002 Date: January 10, 2014 To: All Volatility Index Derivatives Market Participants From: Regulatory Services Division Research and Product Development Department RE: Modified HOSS Opening Procedures for All Volatility Index Option Components CBOE Futures Exchange, LLC (CFE) lists futures (including security futures) on different volatility indexes that are calculated using option series traded on the Chicago Board Options Exchange, Incorporated (CBOE). These option...»

«VI. Resources for Trainers Introduction This annotated list of online, print, and multimedia resources provides professional developers with both seminal professional development resources (e.g., California Adult Literacy Professional Development Project [CALPRO] and National Center for the Study of Adult Learning and Literacy [NCSALL] Study Circle guides) and materials specifically about teaching adult ESL (e.g., Public Broadcasting System [PBS] ESL/Civics Link and Teaching Adult Second...»

«MEMOIRS OF THE GEOLOGICAL SOCIETY OF FRANCE Published with the assistance of the NATIONAL CENTER FOR SCIENTIFIC RESEARCH NEW SERIES — VOLUME XXXIX — ISSUES 1 TO 6 — PLATES I to XVII MEMOIRS No. 88A, P. 1 TO 57, PL. I TO XI 88B, P. 1 TO 34, PL. I TO VI No. THE DINOSAURS OF THE “CONTINENTAL INTERCALAIRE” OF THE CENTRAL SAHARA BY A. F. DE LAPPARENT STUDY OF SOME VERTEBRATES FROM THE MUSCHELKALK OF DJEBBEL REHACH (SOUTHERN TUNISIA) BY F. GORCE PARIS AT THE SEAT OF THE GEOLOGICAL SOCIETY...»





 
<<  HOME   |    CONTACTS
2016 www.thesis.xlibx.info - Thesis, documentation, books

Materials of this site are available for review, all rights belong to their respective owners.
If you do not agree with the fact that your material is placed on this site, please, email us, we will within 1-2 business days delete him.