Welcome to Satoshi Suzuki's Home Page
Japanese version is here.
Profile
1967 Born in Tokyo.
1990/3 Graduated from University of Tokyo
1990/4 Entered NTT
1992/10 Joined ATR Human Information Processing Research
Laboratories
1997/3 Back to NTT Communication Science Laboratories
Research Interests
Computational approaches to brain research, especially, visual recognition, natural nanguage processing based on stochastic modeling.
KEYWORDS:
object recognition, classification, unsupervised learning, competitive
learning, information recovery, modular structures, bi-directional neural
connection, self-organization, feature columns, hierarchical network, etc.
Publishes
-
S.Suzuki, "Probabilistic Word Vector and Similarity based on
Dictionaries," Computational Linguistics and Intelligent Text Processing
Lecture Notes in Computer Science N 2588, Springer-Verlag, 2003
(Proceedings of CICLing-2003) pp.564-574, 2003.
(PS(700K),PDF(200K))
-
S.Suzuki and N.Ueda, "Adaptive Clustering Using Modular Learning
Architecture," Systems and Computers in Japan, vol.34, no.2, 2003.
-
S.Suzuki, "Probabilistic Word Vector based on Dictionaries," Information
Technology Letters(FIT2002), vol.1, pp.79-80, 2002. (in Japanese)
(PS(220K),PDF(330K))
-
S.Suzuki and N.Ueda, "SMEM algorithm for mixture regression problems,"
Electronics and Communications in Japan, Part 2, vol.84, no.12, 2001.
-
S.Suzuki and N.Ueda, "SMEM algorithm for mixture regression models", IEICE D-II, vol.J83-D-II, no.12, pp.2777-2785, 2000. (in Japanese)
(PDF(5.1M))
-
S.Suzuki and N.Ueda, " Adaptive clustering method using modular learning architecture", IEICE D-II, vol.J83-DII No.6, pp.1529-1538, 2000. (in Japanese)
(PDF(497K))
-
S.Suzuki and H.Ando, "A modular network scheme for unsupervised
3D object recognition", Neurocomputing, vol.31, pp15-28, 2000.
(PDF(381K))
-
S.Suzuki and N.Ueda, "Learning NGnets by SMEM algorithm", JNNS'99, Sapporo, pp145-146, 1999. (in Japanese)
(PDF(146K))
-
H.Ando, S.Suzuki and T.Fujita, "Unsupervised visual learning of three-dimensional objects using a modular network architecture", Neural Networks, vol.12, pp1037-1052, 1999.
-
S.Suzuki and N.Ueda, "Adaptive clustering method using modular learning architecture", Technical Report of IEICE,
NC98-181, pp.215-222, 1999. (in Japanese)
-
S.Suzuki and N.Ueda, "Self-Organization of Feature Columns and its Application
to Object Classification", ICONIP'97, Springer, pp.1166-1169, 1997.
(PDF(157K))
-
S.Suzuki and N.Ueda, "A Model for Self-Organization of Feature Columns
using Modular Structure", JNNS'97, Kanazawa, pp133-134, 1997. (in Japanese)
-
S.Suzuki and H.Ando, "3D Object Recognition from 2D Views", IEICE D-II,
vol.J79-DII No.7, pp.1291-1300, 1996. (in Japanese)
-
S.Suzuki and H.Ando, "Unsupervised Classification of 3D Objects from 2D
Views", NIPS 7, The MIT Press, Cambridge,MA., pp.949-956, 1995.
(Abstract)
(PDF(874K))
-
S.Suzuki and H.Ando, "Recognition and Classification of 3D Objects Using
a Modular Learning Network", Technical Report of IEICE, NC93-62, 1993.
(in Japanese)
Awards
1994 JNNS (Japanese Neural Network Society) Research Prize "Recognition
and Classification of 3D Objects Using a Modular Learning Network"
Access
Satoshi Suzuki
NTT Communication Science Laboratories
Address: 2-4, Hikaridai, Seika-cho, Soraku-gun, Kyoto 619-0237 Japan
phone: + 81 774 93 5371
facsimile: + 81 774 93 5345
s u z u k i . s @ l a b . n t t . c o . j p
Last modified on: Mar 27 2003