Usigome, Tokyo, JAPAN, 1938
| 1961 | Graduation from the University of Tokyo, Faculty of Engineering |
| 1963 | Graduation from Postgraduate School, the University of Tokyo |
| 1975 | Ph.D. from the University of Tokyo, Mathematical Engineering |
| 1963-1989 | Central Research Laboratory, Hitachi Ltd. Engaged in Speech Recognition, Character Recognition, etc. |
| 1989-2003 | Professor of Shinshu University, Faculty of Engineering |
| 2003-2007 | Professor of Kyushu Sangyo University, Faculty of Information Science |
IEEE, IEICE (Japan), ISJ (Lifetime Member), JSAI, and AVIRG
In 1960's, PCM (Pulse Code Modulation) was going into practical use for telephone transmission. Since Hitachi had failed to enter transmission business at that time, the "PCM" project started aiming at the entry into the area. The main purpose of the project was development of equipments, of course. The Acoustics Laboratory was ordered to cooperate and help the main team. Since Hitachi's system was slightly different from the NTT's, Prof. Nakano and the members were afraid that NTT's criteria might not fit to the system, so they decided to work on the transmission quality measure by themselves. For the purpose, they developed an experimental one channel PCM system, whose parameters could easily be controlled. By changing its design parameters, such as bit number, error probability, filter bandwidths, etc., they measured the detectability of noises caused by incompleteness of PCM systems.
In 1960's, formant synthesis systems were dominant for speech synthesis. The research group including Prof. Nakano challenged a vocal-tract analog synthesizer in cooperation with ETL (Electro-Technical Laboratory) and developed a pilot model consisting of 60 operation amplifiers! Of course, IC was not invented in the period, so an operation amplifier was assembled on a package using transistors. Frequency response of the amplifier was not good, so they should have fought with the system oscillation every hour.
Prof. Nakano joined a developing team aimed for speech synthesis, who succeeded to develop an Interactive Conversation System via telephone with synthetic speech responses. The Hitachi system has been utilized in the Japanese Railway's Shinkansen (bullet train) ticket reservation system since early 1970's.
DP matching technique (sometimes called as DTW) proposed by Dr. H. Sakoe (formerly with NEC, now a professor emeritus at Kyushu University) shocked speech researchers worldwide. Prof. Nakano followed this technique and recognized its importance. Their paper pointed out the role of feature sets used with this technique.
Prof. Nakano's main concern was the improvement of target detection performance using passive sonar. He was one of the pioneers who utilized FFT for acoustic signal processing in Japan.
Since it was important to show the analyzed sound spectrum to human operators in an enhanced manner, they proposed a colored display system for the spectrogram.
Prof. Nakano proposed a method using peripheral distributions for printed Kanji recognition. To tell the truth, the proposed technique itself has not gone into the practical usage. However, his technique very much influenced researches of OCR in Japan, especially for handwritten Kanji character recognition.
From 1977 to 1989, Prof. Nakano engaged in the development of handwritten character recognition projects. The target vocabulary included numerals, alphabets and katakana (ANK) in the former half of the period, while it covered handwritten kanji at the latter half. The developed algorithms were installed into the Hitachi's OCR. Since the numerals to be read by the OCR's were mainly used in accounting forms in the business, the requirement to make substitution errors negligibly small was very severe. The main efforts were devoted in order to make substitution errors minimal. The ANK OCR development project was assigned to be classified, so they could not write any papers on the achievements. After project finished, the research looked very obsolete for them, because several years had already elapsed after the peak of the development. Prof. Nakano has noticed recently no papers were published regarding the project which consumed the labor of some hundred man-months. He was very sorry for his colleagues on the infertile academic achievements of the project. As for the kanji recognition, it is difficult to recognize a sole character by itself; many pairs cannot be distinguished even by the human beings. So the AI approach, or post-processing, plays a very important role in kanji recognition. His effort was related to the high speed post-processing of recognized results. The work did not end in his Hitachi days, but his young colleagues have made many splendid works after he resigned.
In Prof. Nakano's opinion, the segmentation of characters is an essential task in character recognition. This opinion may be easily accepted by U.S. and European researchers. He felt, however, many researchers in Japan are engaged in the recognition algorithm of the isolated characters and did not pay attention to the segmentation problem, esp. for the cursive English words. We think he has developed, FIRST IN THE WORLD, a practical system which could recognize the touching handwritten numerals.
Prof. Nakano and his group developed a document image understanding system incorporating with character recognition. We think that this was the first system in the world which could work in the practical environment. Though the system could handle the table form documents only, it was really the first operable machine and they exhibited it in 1984.
To make the document understanding universal, the usage of a formal language description for the knowledge on form. They first realized such a system. It is regrettable that many researchers believe FDL is only applicable to fixed forms because FDL describes the documents using absolute values. This is misunderstanding. FDL adopts a generic representation of documents.
Prof. Nakano was interested in facial image detection, esp. one using textures. Recently, they developed a neural-network-based algorithm for face recognition.
(This information was edited mainly from the HP of Kyushu Sangyo University)