item3

What's New | Greetings | Schedule | Members | Contact | Japanese

After March 2020

2020

March

February

January

2019

December

November

October

September

August

July

June

May

April

March

February

January

2018

December

November

October

September

August

July

June

May

April

March

February

January

2017

December

November

October

September

August

July

June

May

April

March

February

January

2016

December

November

October

September

August

July

June

May

April

Before April 2016

Schedule in August 2016


The 28th Perceptual Frontier Seminar: Speech and Music

Date and time: Wednesday, 10 August 2016, 16:30-18:00
Venue: 411, 1st floor, Build. 4, Ohashi Campus, Kyushu University
Organizer: Kazuo UEDA (Kyushu University/ReCAPS)

The Research Center cordially invites you to participate and make a presentation at the Seminar. We will get together after the Seminar for a reception party of Ms. Katharina ROST and dinner.

Program

1. "Pixelated speech": The role of spectral and loudness changes for the irrelevant speech effect
Katharina ROST*, Josef SCHLITTENLACHER*, and Wolfgang ELLERMEIER*
*Technical University of Darmstadt

The “irrelevant speech effect” describes the phenomenon that irrelevant background speech substantially disrupts memory performance; it is not sufficiently understood yet. The present experiment explored to what extent spectral and loudness changes influence the magnitude of irrelevant speech effect using "pixelated speech". Spectral changes seem to be the dominant factor in accounting irrelevant speech effect but its disrupting effect is surpassed with a combination of spectral and loudness changes.

2. The perception of a dotted rhythm embedded in a two-four-time framework
Chinami ONISHI*, Yoshitaka NAKAJIMA**, and Emi HASUO***
*Human Science Course, Kyushu University, **Department of Human Science/Research Center for Applied Perceptual Science, Kyushu University, ***College of Education, Psychology and Human Studies, Aoyama Gakuin University

The purpose of our study was to examine how a beat framework would affect discrepancy observed between proportion of physical durations and musical notation in a dotted rhythm. The participants were instructed to adjust the physical ratio between adjacent durations so as to reflect a notated rhythm as precisely as possible. The results indicated a tendency that the adjusted physical proportions of a dotted quaver were larger than a notated proportion, and this tendency was more prominent when a semiquaver preceded the dotted quaver, except when the dotted rhythm was located on the first beat of the second bar.

3. The temporal structure of English speech in public speaking
Shimeng LIU*, Yuko YAMASHITA**, Yoshitaka NAKAJIMA*, Kazuo UEDA*, Gerard B. REMIJN *, and Miharu FUYUNO*
*Kyushu University, **Shibaura Institute of Technology

To investigate the role of pause duration in public speaking, we recorded 11 Japanese university students' speech in Fukuoka during an English speech contest. The pause durations in speech were semi-automatically extracted and measured by using a computer software praat; the results showed that the pause duration for periods had to be longer than for commas and others, to achieve high scores at the contest. We are planning to obtain more English speech samples spoken by non-natives with other language backgrounds.

Photos


The 29th Perceptual Frontier Seminar: Poster Symposium in Summer

Date and time: Monday, 22 August 2016. Part I, 15:00-15:50; Part II, 16:00-18:00
Venue: Part I, Office of Kazuo UEDA (709), 7th Floor, Build. 3, Ohashi Campus, Kyushu University; Part II, Lobby, 1st Floor, Build. 3, Ohashi Campus, Kyushu University
Organizer: Yoshitaka NAKAJIMA (Kyushu University/ReCAPS)

We will get together to have a small party in 321 after the Seminar.

Most of the posters are supposed to be exhibited in the lobby of the 5th Building from 29 August to 9 September 2016.

Program

Part I: Oral Presentation

15:00-15:50

1. Audiovisual correspondences between cover art images and music tracks
Chihiro HIRAMATSU*, Shinnosuke OGATA*, and Takeharu SENO*
*Kyushu University

Factors influencing audiovisual correspondences were investigated using cover art images associated with music tracks. Music-image matching, visual classification of cover art images and statistics of image and music features were analyzed. Cognitive and/or semantic aspects of images and music features related to rhythm, tone and timber were suggested to be important factors.

Part II: Poster Symposium (with posters for other conferences)

16:00-18:00

1. Making demo-movies of various vection stimuli
Emi SETOGUCHI* and Takeharu SENO*
*Kyushu University

Making demo-movies of vection stimuli that were based on over 30 psychological published articles, as many as possible was challenged. There has been no such challenge in psychology so far. Those movies can contribute to science and also art-entertainment.

2. Effects of prior walking context on the vection induced by different types of global optic flow
Takeharu SENO*, Stephen PALMISANO**, and Shinji NAKAMURA***
*Kyushu University, **University of Wollongong, ***Nihon Fukushi University

In this study, the effects of prior walking context on the vection were investigated. Nineteen participants walked or sat for five minutes while wearing either iron or wooden clogs, and then vection was induced by the radial flow with or without added simulated viewpoint oscillation. The results showed that vection strength was inhibited by walking with iron clogs (relative to walking with wooden clogs or being seated with both types of clogs) and always facilitated by adding oscillation.

3. A new bias in repeated serial subjective estimation
Emi SETOGUCHI* and Takeharu SENO*
*Kyushu University

The results of 20 contests of Japanese comedians (R-1, M-1 and King of Conte) in which 8 comedians were repeatedly estimated by their skills of comedy by professional reviewers, was analyzed. Secondary, a vection experiment in which 12 naïve participants subjectively estimated vection strength 8 times was conducted. Then the obtained values became increased along with the trial numbers.

4. Construction of a model of vection
Takeharu SENO*, Ken-ichi SAWAI**, Masaki OGAWA*, Toshihiro WAKEBE+, Hidetoshi KANAYA++, Kim JUNO+*, and Stephen PALMISANO***
*Kyushu University, **University of Tokyo, +Fukuoka Jo Gakuin University, ++Aichi Shukutoku University, +*University of Sydney, ***University of Wollongong

A model of vection by using three indexes (latency, duration and magnitude of vection) was presented in this study. The model was consisted of two waves that vary over time, i.e. sinusoidally modulated strength and exponentially increased strength and we also simulated 10000-trials by using our model. Then the simulated results of the distributions of latency, duration and magnitude were very similar to the obtained real data.

5. Subjective sleepiness and vection strength can be correlated
Takeharu SENO*, Masaki OGAWA*, Kouhei MATSUMORI*, and Shigekazu HIGUCHI*
*Kyushu University

This study revealed that the sleep deprivation had a facilitation effect on self-motion perception (vection). The strength of vection was measured in two conditions: sleep-deprivation and normal sleep conditions. In the sleep deprivation condition, participants did not sleep for about 20 hours. Magnitudes of vection and subjective sleepiness were significantly positively correlated.

6. Developing effective instructions to decrease Japanese speaker's nervousness during English and Japanese public speeches: Evidence from psychological and physiological measurements
Miharu FUYUNO*, Yuko YAMASHITA**, Yuki YAMADA*, and Yoshitaka NAKAJIMA*
*Kyushu University, **Shibaura Institute of Technology

This study examines the nervousness of Japanese speakers during public speeches in English and Japanese to develop an instruction module that is effective for decreasing speakers' nervousness. Heart-rate variability values, Profile of Mood States (POMS2), and original questionnaire sheets were used to evaluate the nervousness in participants. The results indicated that the experimental group showed less nervousness, especially during English speeches, in the second performance.

7. Development of a multimedia Spanish language learning application: based on ADDIE model and Unity 3D platform
Laura Maria BLANCO CORTES* and Miharu FUYUNO*
*Kyushu University

This study aims to develop a multimedia Spanish language learning application with a focus on authentic materials and ADDIE (Analysis-Design-Development-Implementation-Evaluation) model. After analyzing needs and designing lesson contents, a prototype of application was developed with Unity 3D engine. This study finally discusses results from a questionnaire survey that was performed to evaluate the effects of the prototype.

8. The effect of continuity of motion on vection
Yoshitaka FUJII* and Takeharu SENO*
*Kyushu University

The effect of frame rate of movie on vection was investigated in wide range (3 to 60 fps). Downward and expanding grating movies were used as vection stimuli and the strength of vection was measured. The results showed that a significant effect of frame rate was obtained in both stimuli, and vection induced by expanding grating had more rapid increase with increasing the frame rates than by downward grating.

9. The effect of frame rate and motion blur on vection
Yoshitaka FUJII*1*2, Robert S. ALLISON*1, Pearl S. GUTERMAN*1, and Laurie M. WILCOX*1
*1 York University (Canada), *2 Currently in Kyushu University

The effect of frame rate and motion blur on vection was investigated in this study. Observers moved forward in a town rendered with 3-D computer graphics and responded vection strength. The results show that motion speed has very strong effect, frame rates also have significant effect, but motion blur has no effect to vection.

10. Effect of body posture on the interpretation of cast shadows
Tomomi KOIZUMI*, Hiroyuki ITO*, Shoji SUNAGA*, and Erika TOMIMATSU*
*Kyushu University

It was investigated whether the light-from-above (or above-left) assumption was used in the interpretation of cast shadows, and how gravity influenced this light assumption. With four observation postures (sitting, lying-on-left, right, and back), relative strength of perceptual correspondence between the center disk and the adjacent cast shadows was measured. Results indicated that the light-from-above-left assumption was used and the directional bias of preferred shadow might have been determined by both body posture and gravity.

11. Importance of trichromatic color vision for detecting monkey facial coloration: A functional substitution approach
Chihiro HIRAMATSU*, Amanda D. MELIN** William L. ALLEN***, Constance DUBUC****, and James P. HIGHAM*****
*Kyushu University, **University of Calgary, ***Swansea University, ****University of Cambridge, *****New York University

Optimal color vision to detect reproductive status of female macaque monkeys was investigated. Performance in detecting a proceptive (i.e. sexually active) face against a non-proceptive face was compared among five simulated color vision types and common trichromatic color vision. Results suggested that common trichromatic color vision confers excellent ability to detect meaningful variation in primate face color.

12. The effect of sound on visual grouping in a multi-stable stimulus
Hiroaki YANO*, Yoshitaka NAKAJIMA **, Kazuo UEDA**, and Gerard B.REMIJN**
*Graduate School of Design, Kyushu University, **Department of Human Science/Research Center for Applied Perceptual Science, Kyushu University

The effects of auditory stimuli on Polka dance perception, which reflects audiovisual integration on visual grouping of moving objects, was investigated. The task of the participants was to report the motion percept of the two bars on the computer screen, as perceived during the final motion of the bars. The results showed that the presentation of a click at the bars’ point of coincidence facilitated the bouncing percept, and auditory motion information from a click train panned between the speakers facilitated visual clockwise rotation.

13. Cortical hemodynamic response patterns to normal and whispered speech in 3- to 7-year-old children
Gerard B. REMIJN*, Mitsuru KIKUCHI**, Yuko YOSHIMURA**, Kiyomi SHITAMICHI**, Sanae UENO**, and Yoshio MINABE**
*Faculty of Design/Research Center for Applied Perceptual Science, Kyushu University, **Faculty of Medical Science/Research Center for Child Mental Development, Kanazawa University

The purpose of this study was to assess cortical hemodynamic response patterns in 3- to 7-years old children listening to two speech modes: normally-vocalized and whispered speech. Near-infrared spectroscopy (NIRS) was used to assess changes in cortical oxygenated hemoglobin from 16 typically developing children. Specifically over areas that represent speech articulatory body parts and motion, such as the larynx, lips, and jaw, the children showed higher oxygenated hemoglobin for whisper than for normal speech over left ventral sensory-motor cortex (vSMC), which can be attributed to inner (covert) practice or imagination of the different articulatory actions necessary to produce whisper as opposed to normal speech.

14. Effects of factor elimination on the intelligibility of noise-vocoded Japanese speech
Takuya KISHIDA*, Yoshitaka NAKAJIMA**, Kazuo UEDA**, and Gerard B. REMIJN**
*Graduate School of Design, Kyushu University, **Department of Human Science/Research Center for Applied Perceptual Science, Kyushu University

The purpose of the study was to investigate perceptual roles of 4 power-fluctuation factors extracted from 20 critical-band power fluctuations of speech with an origin-shifted principal component analysis followed by a varimax rotation.  Three male native speakers of Japanese listened to noise-vocoded Japanese speech resynthesized from 3 of the 4 power-fluctuation factors, and reported words what they heard.  Intelligibility of the resynthesized noise-vocoded speech considerably decreased when factor 2, covering 570-1370 Hz, was eliminated.

15. Origin-shifted principal component analysis: A method suitable for reconstructing non-negative data
Takuya KISHIDA*, Kanshi NAKAO*, and Yoshitaka NAKAJIMA**
*Graduate School of Design, Kyushu University, **Department of Human Science/Research Center for Applied Perceptual Science, Kyushu University

Power fluctuations of speech sounds separated into 20 critical bands were analyzed with a modified version of principal component analysis, i.e., origin-shifted principal component analysis, in which calculated eigenvectors originated from the zero point of the data.  The extracted principal components were suitable for resynthesizing speech sounds because the resynthesized speech sounds were not accompanied by stationary background noise as would have appeared in a conventional principal component analysis as the error in predicting the acoustically silent point.  Origin-shifted principal component analysis seems to be applicable to analyze and resynthesize various non-negative multivariate data. 

16. The speech to song illusion in Japanese
Asuka ONO*, Yoshitaka NAKAJIMA**, Kazuo UEDA**, and Gerard B. REMIJN**
*Graduate School of Design, Kyushu University, **Department of Human Science/Research Center for Applied Perceptual Science, Kyushu University

A spoken sentence can be heard as a sung sentence by simply repeating it. So far, this so-called speech-to-song illusion has been investigated with stress-accented, stress-timed languages with regular durational contrasts between vowels, with the use of a 5-point rating scale ('1' for 'speech', '5' for 'song'). Here we show that the illusion also occurs in Japanese, a mora-timed, pitch-accented language with a regular duration of morae and strict rules on pitch patterns.

17. Change, not motion, determines subjective duration
Erika TOMIMATSU*, **, Yoshitaka NAKAJIMA*, Mark A. ELLIOTT*,***,**** and Hiroyuki ITO*
*Kyushu University, ** Japan Society for the Promotion of Science, ***National University of Ireland, Galway, ****Sunway University

The purpose of the present study was to reveal the influence of temporal and spatial changes on perceived durations. We inserted some blanks into dynamic-random-dot and static-random-dot durations in order to compare the perceived durations with/without the temporal and spatial changes, and participants answered the perceived durations by the method of reproduction. Our results show that the temporal changes were enough to lengthen the perceived duration.

18. Intelligibility of locally time-reversed speech in four languages: Chinese, English, German, and Japanese
Kazuo UEDA*, Yoshitaka NAKAJIMA*, Wolfgang ELLERMEIER**, and Florian KATTNER**
*Department of Human Science/Research Center for Applied Perceptual Science, Kyushu University, **Institut fuer Psychologie, Technische Universitaet Darmstadt

It was examined how the intelligibility of spoken sentences in Chinese, English, German, and Japanese changes as the duration of segments locally reversed in time increases.  Intelligibility was above 90% when segment duration was 45 ms, but dropped with increasing segment duration to below 10% at 120 ms.  The 50% intelligibility point was observed for segment durations of around 65-80 ms irrespective of language.  This finding and the majority of the previous literature strongly suggest that a common time-averaging mechanism works in speech perception across different languages.

Photos

What's New | Greetings | Schedule | Members | Contact | Japanese