 |
 |
 |
Professor |
Associate Professor |
Assistant Professor |
- Implementation and evaluation of similar segment search algorithm in an arbitrary number of time series
- Theoretical analysis and evaluation of distance-based Active search algorithm
- Voice interval detection in noisy environments
- Participation to Open Campus with lab. demonstrations. Promotion of a series of computer training courses for visual handicapped people in Aizu region.
- Computational Auditory Scene Analysis
There was intensive research on the integration and segregation factors for sound component self organization. We conducted some examinations on the interactions between different primary cues for sound integration and segregation in human auditory systems by psychological experiments and tried to find out quantitative relations for computational auditory scene analysis.
- Robotic Spatial Sound Localization and Sound-based Position Identification
Audition is important for mobile robots as well as vision and other sensing systems. We developed a robotic spatial sound localization system using an auditory interface with four microphones arranged on the surface of a spherical robot head. The time difference and intensity difference from a sound source to different microphones are analyzed and used as localization cues. On the other hand, if a mobile robot can localize several sound sources at known position, the robot will be able to find out its own position. Comparing to vision based systems, the sound based position identification method has advantages in poor lighting conditions and robust against obstacles.
- Research on 3-D Sound Systems
- Speaker based 3-D sound systems
- Analysis of Head Related Transfer Functions
- Reverberation and reality in 3-D sound systems
|
Human Interface Lab. is consisted of four faculty members. Each member has own interest and are doing independent research activities; Prof. Masahide Sugiyama: Prof. Jie Huang: |
Refereed Journal Papers
| [j-huang-01:2006] |
S. Ding, J. Huang, D. Wei, and A. Cichocki. A Recursive Blind
Source Separation Algorithm for Real-Time Processingof Acoustic Signals.
IEEE Transactions on Circuits and Systems I, 53(1):114-128,
2006. |
| [sugiyama-01:2006] |
M. Sugiyama and T. Okamoto. An Efficient Segment Search
Method Based upon Geometric Properties of Metric Space and Output
Probability Sequence. Trans. of IPSJ, 47(6):1675-1686, June 2006. |
| Several studies on efficient segment search methods have been carriedout.
This paper proposes a new efficient segment searching methodbased on
geometric properties of metric space and output probabilitysequence and
evaluates the effectiveness of the proposed methodthrough searching experiments.
The new search method is based on theproperty of triangle inequality
in mathematical distance. Theefficient searching is implemented based
on the following threecomponents; determination using segmentation centers
(M1),determination using clustering centers (M2), and distance pruning(
M3). When the determination can not be performed using the abovethree
components, the conventional Active search (AS) is applied. Theexperimental
results show that the proposed method (M1+M3+AS) runsabout 12 .20
times less distance computation and 26 . 36 timesfaster than the conventional
Active searching, M1+M3 is more efficientabout 37 times less distance
computation and 5 times faster thanM1+M2, and a combination of M1+M3
and binary searching achieves muchfaster. |
|
| [sugiyama-02:2006] |
S. Takeuchi and M. Sugiyama. Evaluation of Automatic Model
Training Method for Voice/Non-Voice Classification. Trans. of IPSJ,
47(6):1774-1781, June 2006. |
| The purpose of this paper is to detect voice section of these multimediacontents.
By removing unnecessary section (noise, music, and so on)from sound
data, speech recognition techniques can prevent missrecognition. This section
detection technique can use pre-process ofspeech recognition techniques. This
paper investigates voice/non-voicediscrimination method and its automatic
training method. By creatingteaching signal from target data, it is able to
apply recognitiontechniques to some data that it is difficult to do previous
training.Also it is able to prevent reduction of effectiveness by the differenceof training data and testing data. As a result, proposed method shows4.2% of
classification error rate. The efficiency of voice/non-voicemodel which is created
by proposed method equals to model which iscreated by about 40% to
50% of manually created teaching signaland proposed method can reduce
manual creating work of teaching signal. |
Refereed Proceeding Papers
| [j-huang-02:2006] |
J. Huang, C. Zhao, Y. Ohtake, H. Li, and Q. Zhao. Robot
Position Identification Using Specially Designed Landmarks. In Proc.
IEEE Instrum. Meas. Technol. Conf. IEEE, 2006. |
| [j-huang-03:2006] |
J. Ootomo, K. Tanno, A. Saji, J. Huang, and W. Hatano. The
Precedence Effect of Sound From Side Direction. In Proc. AES 30th Int.
Conf., 2007. |
| [j-huang-04:2006] |
J. Huang, S. Ishikawa, M. Ebana, H. Li, and Q. Zhao. Robot
Position Identification by Actively Localizing Sound Beacons. In Proc.
IEEE Instrum. Meas. Technol. Conf. IEEE, 2006. |
| [sugiyama-03:2006] |
Masahide Sugiyama. Computational Complexity of a
Distance-Based ActiveSearch Algorithm. In Proc. of 2006 Joint Meeting,
pages -, Nov. 2006. |
| An efficient pattern (segment) search in a large-scale database is avery important
technology in current widespreaded Internet society. Toachieve efficient
searches, many efficient pattern search algorithmshave been proposed. Active
Search (AS) is one of them; it is based onskipping of search calculation. Initially,
AS was formulated betweentwo histograms using a similarity measure.
In fact, AS can beformulated using L p distance, where p is greater than 1.
The searchefficiency can be evaluated using a number of distancecalculations.
The fist result in this paper shows that the L 1distance in AS gives the minimum
number of distance calculations amongL p distances. This property can
be derived from a newly derivedinequality between the L p and L 1 norm on
output probability space. Anew distance can be derived from positive weighting
of anydistance. The second result of this paper is that the average skipwidth for
any weighed L p distance can not be greater than the mostefficient distance.Therefore, no positive combination of L p can bemore efficient than L 1. The
two results indicate that L 1 distance isthe most efficient for Active Searching. |
|
| [sugiyama-04:2006] |
Masahide Sugiyama. A New Similar Segment Search Algorithm
in MultipleTime Series. In Proc. of MMSP2006, pages -, Oct.
2006. |
| This paper formulates a search problem of similar/common segments inmultiple
time series and proposes a new recursive search algorithm;Recursive Diamond
Division Search (RDDS), and shows the effectivenessof RDDS for the similar
segment search problem in two time series anda query search problem in one
time series. The experimental resultsshows that RDDS in two time series runs
the same as or faster than theconventional RIFAS, and RDDS in query search
problem in one timeseries runs faster than the conventional Active Search.. |
Unrefereed Papers
| [sugiyama-05:2006] |
Masahide Sugiyama. On Ball Covering Problem in Similar
Segment SearchAlgorithm. In Proc. of IPSJ Tohoku Regional Workshop,
pages 06.2-29. IPSJ, Jan. 2007. |
| [sugiyama-06:2006] |
A. Sakai and M. Sugiyama. An Efficient Algorithm for Similar
Segment Search-Implementation and Evaluation in Three Time Series-.
In Proc. of IPSJ Tohoku Regional Workshop, pages 06-2-30. IPSJ, Jan.
2007. |
| [sugiyama-07:2006] |
Masahide Sugiyama. Music Analysis Based on Similar Patterns.
In Proc. of IPSJ Tohoku Regional Workshop, page 7. IPSJ, Feb.
2007. |
| [sugiyama-08:2006] |
T. Haga and M. Sugiyama. Comparison between Image and
Sound Information inVideo Retrieval. In Proc. of IPSJ, pages 6T-9.
IPSJ, Mar. 2007. |
| [sugiyama-09:2006] |
Masahide Sugiyama. On Computation Complexity of
Distance-based ActiveSearch Algorithm. In Proc. of ASJ Fall Meeting,
pages 2-P-17, Japan, Sep. 2006. ASJ. |
| [sugiyama-10:2006] |
Masahide Sugiyama. On Diamond Covering in Similar Segment
Search RDDSAlgorithm. In Proc. of ASJ Spring Meeting, pages
1-P-22, Japan, Mar. 2007. ASJ. |
| 1-P-22 |
|
| [sugiyama-11:2006] |
S. Takeuchi and M. Sugiyama. Evaluation of unsupervised
voice detection method usingCENSREC-1-C. In Proc. of ASJ Spring
Meeting, pages 91-92. ASJ, Mar. 2007. |
| 3-9-14 |
|
| [sugiyama-12:2006] |
Masahide Sugiyama. On Computation Complexity of
Distance-based Active SearchAlgorithm. In Report of Speech Research
Committee, pages 19-24, Japan, June 2006. IEICE/ASJ. |
| SP2006-9 |
|
| [sugiyama-13:2006] |
Y. Watanabe and M.Sugiyama Y. Tanimoto, A. Sakai. Evaluation
of SimilarSegment Search Algorithm in Multiple Time Series. In
Report of Speech Research Committee, pages 83-88, Japan, Dec. 2006.
IEICE/ASJ. |
| SP2006-99 |
|
| [sugiyama-14:2006] |
K. Watanabe and M. Sugiyama. Audio Document Retrieval
with its Transcription. In Proc. of SDPWS2007, pages 65-72, Feb. 2007. |
| SP2006-99 |
Grants
| [sugiyama-15:2006] |
Masahide Sugiyama. Competitive Research Funding in UoA,
2006-2007. |
Academic Activities
| [sugiyama-16:2006] |
Masahide Sugiyama, 2006. Member of Council, ASJ |
| [sugiyama-17:2006] |
Masahide Sugiyama, 2006. Reviewer of IEICE Transaction |
Patents
| [sugiyama-18:2006] |
M. Sugiyama. Similar/non-similar determination machine,
method and its program, Oct. 2006. |
Ph.D and Other Theses
| [j-huang-05:2006] |
Yuka Watanabe. Graduation Thesis: A 3-D Sound System
Using 8 Speakers Arranged on the Horizontal Plane, University of Aizu,
2006. Thesis Advisor: Huang, J. |
| [j-huang-06:2006] |
Koga Seiya. Graduation Thesis: Adjustment of the Effect of
HRTFs for Speaker based 3-D Sound Systems, University of Aizu, 2006. Thesis Advisor: Huang, J. |
| [j-huang-07:2006] |
Atsushi Tokuda. Graduation Thesis: Listening Position Effect
on a Speaker Based 3-D Sound System, University of Aizu, 2006. Thesis Advisor: Huang, J. |
| [j-huang-08:2006] |
Yumi Kudou. Graduation Thesis: Environmental Sound Analysis
and Synthesis for Computational Generation Using the Wigner Distribution
with Reassignment, University of Aizu, 2006. Thesis Advisor: Huang, J. |
| [j-huang-09:2006] |
Mayuko Ono. Graduation Thesis: Robot Position Identification
by Sound Beacons Located in Know Positions, University of Aizu, 2006. Thesis Advisor: Huang, J. |
| [j-huang-10:2006] |
Khelchandra Thongam. Master Thesis: Motion planning of
a two-joint manipulator in 3-Dimensional environment using sampling
based method, University of Aizu, 2006. Thesis Advisor: Huang, J. |
| [j-huang-11:2006] |
Takuya Yoshiara. Master Thesis: Spatial Sound Localization
Using a Spherical Robot Head, University of Aizu, 2006. Thesis Advisor: Huang, J. |
| [sugiyama-19:2006] |
Yuki Tanimoto. Graduation Thesis: Improvement of A Similar
Segment Search Algorithm, University of Aizu, 2006. Thesis Advisor: M. Sugiyama, Referee: Prof. T. Hayashi |
| [sugiyama-20:2006] |
Akihiro Sakai. Graduation Thesis: An Efficient Algorithm for
Similar Segment Search.Implementation and Evaluation in Three Time
Series-, University of Aizu, 2006. Thesis Advisor: M. Sugiyama, Referee: Prof. R. Oka |
| [sugiyama-21:2006] |
Mihoko Endou. Graduation Thesis: Evaluation of Speaker
Segment Retrieval, University of Aizu, 2006. Thesis Advisor: M. Sugiyama, Referee: Prof.J. Huang |
| [sugiyama-22:2006] |
Masatoshi Sawai. Graduation Thesis: Development of GUIbased
Applications of Doublet-- From Lewis Carrollsf Mind to a Computer
Screen --, University of Aizu, 2005. Thesis Advisor: M. Sugiyama, Referee: Prof. S. Ohkawa |
| [sugiyama-23:2006] |
Teppei Haga. Master Thesis: A Quick Video Retrieval Algorithm,
University of Aizu, 2006. Thesis Advisor: M. Sugiyama, Referees: Profs. R. Oka, and Q. Zhao |
| [sugiyama-24:2006] |
Yuuki Chiba. Master Thesis: 3D Gesture Agent, University
of Aizu, 2006. Thesis Advisor: Masahide Sugiyama, Referees: Profs. T. Hayashi and M.Cohen |
| [sugiyama-25:2006] |
Atushi Idogawa. Master Thesis: Development of Multilingual
Emacspeak, University of Aizu, 2006. Thesis Advisor: M. Sugiyama, Referees: Profs. M. Cohen and J. Huang |
| [sugiyama-26:2006] |
Yoshiyuki Watanabe. Graduation Thesis: A New Query
Search Algorithm for Multimedia Data, University of Aizu, 2006. Thesis Advisor: M. Sugiyama, Referee: Prof. Q. Zhao |
| [sugiyama-27:2006] |
Eitarou Miura. Master Thesis: Personal Information Retrieval
and Display in Video, University of Aizu, 2006. Thesis Advisor: M. Sugiyama, Referees: Profs. R. Oka and J.P. Shin |
