MENU
  • School of the Future
  • Lifestyle of the Future
  • Question of the Future
  • Safari of the Future
  • Talent of the Future
  • Contact
Language
  • English
  • 日本語
Takegawa Lab. - xFuture Lab.
Takegawa Lab. - xFuture Lab.
  • School of the Future
  • Lifestyle of the Future
  • Question of the Future
  • Safari of the Future
  • Talent of the Future
  • Contact
  • English
  • 日本語
  2. Question of the Future

Question of the Future

Greetings


Smartphones, apps, SNS and so on are having a huge impact on the way people think and act. This is accompanied by an increase in social problems such as the danger posed by people walking along looking down at their phones, intense online criticism, fake news, and ‘media illiteracy’. In response to questions such as “How can I overcome communication issues?”, “Why do people feel nervous?”, “What factors are necessary to change people’s thoughts and actions?”, and “What’s a method for conducting effective groupwork online?”, Questions of the Future seeks the essence of human perception (how a person sees things), cognition (how a person understands/explains things), and psychology (how a person feels).

Another change to daily life has been the way the spread of subscription, which allows us to use what we need, when we need it, has transformed how things (including applications and services) are circulated. In subscription businesses, a transaction is not completed simply by selling something; it is necessary to ensure that consumers continue to use that thing. In addition, in a modern society overflowing with ‘things’, it has become difficult to differentiate according to those things alone. This raises questions such as “How can we discover new user markets?”, and “What on earth are the requirements and experiences that those new users are really seeking?”. Questions of the Future uses a UX (User eXperience) research method to clarify these problems.

In Questions of the Future, we use cognitive psychology to delve to the core of the human psyche, and use the UX research method to explore how we can establish a better relationship to people and things.

Prof. Yoshinari Takegawa

Project

(Click thumbnails for details)
If you are using an ad blocker, the content may not be displayed properly

Does changing the face of an online lecturer improve test scores?
An attempt to increase motivation in online lectures
Does consumers’ desire to purchase increase when teleshopping products are introduced by an avatar?
Boost sales with an online avatar!
Does changing an interviewer’s face stop interviewees feeling nervous?
A digital mask to control interviewees’ degree of nervousness
Does the nudge effect increase the number of opinion-exchange comments?
Utilizing peer pressure positively in online groupwork
Does changing the line-up of video with a time-lag make it easier to watch online dance lesson videos?
Lag; it’s the ultimate hindrance to online dance, but can we turn it to our advantage?
Does speeding up or slowing down your clock increase operational efficiency?
Fast-and-slow Clock
Does using ‘shakuri’ affect how good you are at singing?
We got AI to sing! An analysis of what people perceive as good singing
How warping the piano performance data of a pro can help you to understand whether your own performance constitutes good playing?
What are the boundaries of ‘correct’ and ‘incorrect’ playing?
AI can tell you what brass instrument is right for you!?
Like a sorting hat for brass instruments! Supporting a new start for wind music

© Future University Hakodate, Takegawa Laboratory, All Rights Reserved.

練習して上手くなった音ゲースキルをほったらかすとどうなるのか？


Accomplishment
Sasaki, A., Matsuura, M., Matsubara, M., Takegawa, Y., and Hirata, K., “Exploring Patterns of Skill Gain and Loss on Long-term Training and Non-training in Rhythm Game,” In Proceedings of the 16th International Symposium on Computer Music Multidisciplinary Research (CMMR ’23), 1b-3 (Oral) (November 2023).

AI can tell you what brass instrument is right for you!?



Like a sorting hat for brass instruments! Supporting a new start for wind music
Kumiwake Brass-guchi’ (the brass-mouth sorter) is a system in which AI determines the brass instrument that’s just right for you. In wind instrument clubs at middle schools and high schools, members often chose their instruments because they ‘just liked the look of it’ or ‘an upperclassman recommended it.’ However, the instrument that a person is suited to actually differs depending on the shape and thickness of their lips. In this system, AI analyzes the characteristics (shape and thickness) of a person’s lips and proposes the optimum instrument from among the trumpet, trombone, horn and tuba. As a result of making the AI learn from images of the faces of professional players, the system was able to judge the optimum instrument with higher accuracy than human specialists (AI: 56%, human: 28%). Thus, this system can reduce mismatches when choosing instruments, and so in turn reduce worries such as “This instrument is too difficult; I can’t go on with it…” Try out our system and find the instrument that’s right for you!

How warping the piano performance data of a pro can help you to understand whether your own performance constitutes good playing?



What are the boundaries of ‘correct’ and ‘incorrect’ playing?
In expressive fields, such as piano playing and design, the criteria for ‘correct’ and ‘incorrect’ can differ from person to person. In this research, which focuses on differences in piano performance, we aimed to clarify the degree of difference that was recognized as ‘correct’ and the degree beyond which performance was recognized as ‘incorrect’. In an experiment, we prepared audio of an accurate performance by a professional pianist and audio created by intentionally warping the volume and keypress timing of that performance. Furthermore, we created multiple warped audio files, by minutely adjusting the degree of warp. Then, we had 50 subjects listen to the audio files and select whether or not they felt there was something wrong with the performance. As a result, when the volume was warped by 48% or above, the performance was considered to be wrong, whereas when the volume was warped by 16% or under, it was recognized as correct. In addition, when keypress timing was out by 40 milliseconds or more, it was considered to be wrong, while a disparity of 10 milliseconds or fewer tended to be judged correct. This research could be applied to not only music, but also other expressive fields such as design. For example, by creating a gauge to objectively evaluate differences in font size or color scheme, we could reduce discrepancy between the way a design is perceived by its evaluators and creators respectively, thus enabling more efficient improvement of design work.

Accomplishment
Matsui, R., Yanagisawa, Y., Takegawa, Y., and Hirata, K., “Survey on the Auditory Feelings of Strangeness while Listening to Music,” In Proceedings of the 25th HCI International Conference (HCII ’23), pp. 95-–106 (July, 2023).
Matsui, R., Yanagisawa, Y., Takegawa, Y., and Hirata, K., “Investigation of Thresholds for Listening to Piano Performances,” In Proceedings of International Conference on Music Perception and Cognition (ICMPC23), pp. 618–621 (Aug. 2023).

Does speeding up or slowing down your clock increase operational efficiency?



Fast-and-slow Clock

When doing something within a time limit, such as school work or a test, do you find yourself able to concentrate and work efficiently all while keeping time allocation in mind? No doubt many people have experience of devoting a lot of time to each question at the beginning of a test, when there is plenty of time left, only to find themselves running out of time to answer the last questions. Accordingly, in this research, experiment collaborators carried out 10-minutes tasks, during which the clock was sped up for the first 5 minutes, then slowed down for the last 5 minutes, to test whether this increased operational efficiency. We call the clock that runs in this way the Fast-and-slow Clock. In addition, we prepared five different time-scaling factors for the fast-and-slow clock shown to the experiment collaborators (1.1 times, 1.3 times, 1.5 times, 1.7 times, and 1.9 times), and investigated which time-scaling factor best enabled the collaborators to exercise their powers of concentration and tackle the task. As a result, it was found that a time-scaling factor of 1.9 times most increased the collaborators’ performance, while a factor of 1.5 times increased both performance and operational accuracy to a suitable degree and in a well-balanced manner. Evidently, performance does not improve exponentially as the time-scaling factor increases, rather, setting the time-scaling factor that is just right to improve performance to an extent makes it possible to increase your test scores and your concentration. Anyone who finds themselves struggling with time allocation and unable to get to the end of a test should try using the Fast-and-slow Clock!

Does changing the face of an online lecturer improve test scores?



An attempt to increase motivation in online lectures

The development of information technology has led to an increase in lectures held online, enabling participants to attend lectures whenever and wherever they like. However, the problem with online lectures, when compared to face-to-face lectures, is that participants have difficulty maintaining concentration. In this research, we verified whether desire to view the lecture, and concentration, can be increased by enabling online lecture participants to choose an avatar lecturer of which they like the appearance. As a result, it was found that when participants chose avatar lecturers that they liked, they found it easier to understand the lecture content and became less likely to grow bored. In addition, they even obtained better scores on the test that is based on the lecture content. Why don’t you try using your favorite avatar to increase your motivation and test results?

Accomplishment
Kojima, N., Takegawa, Y., Terai, A., and Hirata, K., “Effect of Preference for a Lecturer Disguised as an Avatar on Desire to View Lecturer’s Video,” In International Conference on Human-Agent Interaction (HAI ’22). Association for Computing Machinery, New York, NY, USA, pp. 278–280 (Dec. 2022).

Does consumers’ desire to purchase increase when teleshopping products are introduced by an avatar?



Boost sales with an online avatar!

A character used online as a representation of oneself is called an ‘avatar’, and more and more people are incorporating avatars into their lives as a communication tool. In this research, we verified whether, even when products were actually being introduced by a beginner salesperson, consumers’ degree of trust in the salesperson, and their desire to purchase, increased when they saw products being introduced by an avatar whose appearance they had chosen according to their personal preferences. As a result, it was found that selecting a preferred avatar did lead to an increase in trust and purchase desire. The avatars included a banana character that was not considered appropriate for customer service, yet we were surprised to see that there were some consumers who liked the character and whose purchase desire increased accordingly. By changing a salesperson into an avatar that appeals to each individual consumer, we can expect to see teleshopping products sell better than ever. It is difficult to find a unique avatar that is loved by everyone, but it is easy to have each consumer choose the avatar they like. As it is only online that an avatar’s appearance can be changed so easily, this sales method is can be said to be unique to online shopping.

Accomplishment
Sasaki, M., Noto, K., Takegawa, Y., and Hirata, K., “Practice Progress Visualization System to Support Piano Teachers,” In Proceedings of International Conference on Music Perception and Cognition (ICMPC ’23), pp. 267 (Aug. 2023).

Does changing an interviewer’s face stop interviewees feeling nervous?



A digital mask to control interviewees’ degree of nervousness

Using a digital mask which responds to the movement of a person’s face and alters the expression of an animated character, in the manner of a V-tuber, we transformed an interviewer’s face from ‘original face’ to ‘fierce expression’ to ‘kind face’ and investigated how this affected students’ level of nervousness. As a result, we found that, compared to the original face, using the digital mask reduced the interviewees’ nervousness. Furthermore, we found that the kind face relaxed interviewees’ nerves more than the fierce face. Thus, wearing the digital mask enables an interviewer both to reduce an interviewee’s nervousness by using the avatar with a kind face when taking the interview, and to increase an interviewee’s feeling of nervousness, where appropriate, by using a fierce looking avatar.

Accomplishment
Noguchi, K., Takegawa, Y., Tokuda, Y., Sugiura, Y., Masai, K., and Hirata, K., “Study of Interviewee’s Impression Made by Interviewer Wearing Digital Full-face Mask Display During Recruitment Interview,” Proceedings of the 9th International Conference on Human-Agent Interaction, pp. 323–327 (Nov. 2021).

Does the nudge effect increase the number of opinion-exchange comments?



Utilizing peer pressure positively in online groupwork

An increase in online lectures and groupwork has reduced the opportunities for learners to exchange opinions, and it has become difficult to ‘help each other maintain and increase motivation and improve skills’, which should be the advantage of groupwork. Therefore, in this research we use nudging and verify whether it enables us to maintain motivation in groupwork. Here, nudging means getting people voluntarily to act in a desirable way, without controlling their behavior. Our experiment method involved creating 8 groups of 3 people and having each group create a program and share it with the other groups online. Then, over a period of 5 weeks, we had subjects comment each week on the programs created by the group members. We divided the groups into 3 sets: 3 groups to whom we presented a ‘nudge image’ (a graph of the average wordcount of comments) as a reminder at the end of the week, 3 groups to whom we presented only a written reminder, and 2 groups to whom no reminder was presented. As a result, we found that while the amount of commenting decreased for the groups who did not receive a reminder, for the group who received the nudge image the amount of commenting not only did not decrease from the start, but actually increased. We consider that this result is due to peer pressure, and as such is partly influenced by the specifically Japanese personality trait of not wanting to act differently to other people. Regarding the comments that followed the presentation of the nudge image, a large number of comments offered advice on program coding, or motivational messages. In our next experiment, we will investigate whether commenting can be further increased by presenting a nudge image of a falsified graph that displays wordcount statistics that are actually higher than the average.

Does changing the line-up of video with a time-lag make it easier to watch online dance lesson videos?



Lag; it’s the ultimate hindrance to online dance, but can we turn it to our advantage?

With changes such as the adoption of street dance as an Olympic event, and the introduction of dance as a compulsory class in schools, dance is becoming a familiar part of our lives. Accordingly, there has been in increase in online dance content, meaning that it is now possible to be taught dance by a professional, from the comfort of one’s own home. However, the problem with online dance lessons is that, when there are multiple students, factors such as the Wi-fi connection in each student’s house can cause lag to occur, which makes it impossible for students to keep in time with each other, thus making it difficult for individuals to notice if they are getting it wrong. Therefore, in this research, we investigated whether modifying the video image sent from each student and lining up the images on the screen affects how easy it is to watch the video and to notice when one is out of time. We conducted two experiments, in which the images of students were displayed as a 3x# grid. In the first experiment, when students watched the dance lesson the images were arranged randomly, with no consideration for the difference in lag. In the second experiment, the images were ordered according to lag. After each experiment students answered a questionnaire, the results of which showed that ordering the images made it easy to find images with a small degree of lag and thus to spot when one was out of time. Furthermore, it was found ordering the images also made it easier to look at the screen, and to dance.

Accomplishment
Imai, R., Matsui, R., Yanagisawa, Y., Takegawa, Y., and Hirata, K., “Survey on the Effect of Video Delay in Online Dance with Multiple Participants,” In Proceedings of the 25th HCI International Conference (HCII ’23), pp. 375–-384 (July, 2023).

Does using ‘shakuri’ affect how good you are at singing?



We got AI to sing! An analysis of what people perceive as good singing

When singing is rated in karaoke, the more accurate a singer’s pitch is, the higher the score they receive. However, when humans hear singing, rather than a singer with accurate pitch, it is a singer who incorporates the shakuri technique who is perceived as talented. Shakuri is a singing technique in which the singer, before producing a certain pitch, deliberately produces a lower sound and then returns to the original pitch. Accordingly, in this research, we investigated whether the length and pitch difference of shakuri affects how ‘good’ singing is. Survey participants were presented with two variations of a song incorporating shakuri, and asked to choose which they thought constituted better singing skills. This act was carried out in three sets of six questions. The variations listened to by the survey participants were created using AI, and comprised songs in which the length of shakuri was 20%, 40% and 60% longer than the original sound, and in which the pitch difference was 1 semitone, 2 semitones and 3 semitones lower than the original sound. The survey results showed that it was a length of 20% longer and a pitch difference of one to two semitones lower that were perceived as the best. This shows that it is shakuri with a short length and small pitch difference that sounds best to humans. If you want people to think you’re good at karaoke, rather than trying to be pitch perfect, try introducing a little shakuri.