Oral-History:Sadaoki Furui
About Sadaoki Furui
Born in Japan in 1945, IEEE Life Fellow Sadaoki Furui, (M'79–SM'88–F'93) received the B.S., M.S., and Ph.D. degrees in Mathematical Engineering and Instrumentation Physics from the University of Tokyo, in Tokyo, Japan in 1968, 1970, and 1978, respectively. In 1993, Furui was elevated to IEEE Fellow “for contributions to speech analysis, speech recognition, and speaker identification.” Furui passed away on 31 July 2022 and was an eminent member of the IEEE Signal Processing Society and a leader in the field of speech processing, playing an important role in improving natural communication between humans and machines.
After joining the Nippon Telegraph and Telephone Corporation (NTT) Labs in 1970, he worked on speech analysis, speech recognition, speaker recognition, speech synthesis, speech perception, and multimodal human–computer interaction. From 1978 to 1979, he was a Visiting Researcher at AT&T Bell Laboratories, Murray Hill, New Jersey. He was a Research Fellow and the Director of the Furui Research Laboratory, NTT Labs, and a Professor at the Department of Computer Science, Tokyo Institute of Technology. Most recently he has been President of Toyota Technical Institute at Chicago; Chief Research Director, National Institute of Informatics in Tokyo, Japan; and Professor Emeritus at the Tokyo Institute of Technology, Tokyo, Japan.
Furui authored and co-authored more than 1,000 papers and books mostly in the fields of speech recognition, artificial intelligence, and natural language processing. His book, "Digital Speech Processing, Synthesis, and Recognition" has been published in twenty-six editions between 1985 and 2001). He is best known for investigating human perception of transient sounds during the 1980s, gaining an important understanding of human hearing, Furui provided the first quantifiable measurement of the importance of spectral transition upon intelligibility. By incorporating spectral derivatives, also known as “delta cepstra,” the accuracy of speech recognition systems was greatly improved. Most of today’s practical speech recognition, speaker identification and verification systems incorporate this concept in spontaneous speech.
During his career, Furui’s professional associations include: The Institute of Electrical and Electronics Engineers; the International Speech Communication Association; the Institute of Electronics, Information and Communication Engineers, Japan; and the Acoustical Society of Japan.
Furui received many professional awards and honors including the Paper Award and the Achievement Award from the Institute of Electronics, Information, and Communication Engineers of Japan (IEICE) (1975, 1988, 1993, 2003, 2003, 2008), and the Paper Award from the Acoustical Society of Japan (ASJ) (1985, 1987). He received the Senior Award and Society Award from the IEEE Signal Processing Society (1989, 2006), the International Speech Communication Association (ISCA) Medal for Scientific Achievement (2009), and the IEEE James L. Flanagan Speech and Audio Processing Award (2010). He also received the Achievement Award from the Minister of Science and Technology and the Minister of Education, Japan (1989, 2006), and the Purple Ribbon Medal from the Japanese Emperor (2006). Additional awards and honors include Distinguished Lecturer (IEICE), 1993; Fellow (IEICE), 2001; the Distinguished Achievement Award (IEICE), 2008; Fellow (ISCA), 2008; Medal for Scientific Achievement (ISCA), 2009; NHK Broadcast Cultural Award, 2012; Distinguished Lecturer (ISCA), 2012; Okawa Prize, 2013; Accredited as Person of Cultural Merit by Japanese Government, 2016; Permanent Honorary President of APSIPA, 2020; and ISCA Special Service Medal, 2020.
About the interview
SADAOKI FURUI: An Interview Conducted by Mary Ann Hellrigel, IEEE History Center, 20 December 2021
Interview #872 for the IEEE History Center, The Institute of Electrical and Electronics Engineers, Inc.
Copyright Statement
This manuscript is being made available for research purposes only. All literary rights in the manuscript, including the right to publish, are reserved to the IEEE History Center. No part of the manuscript may be quoted for publication without the written permission of the Director of IEEE History Center.
Request for permission to quote for publication should be addressed to the IEEE History Center Oral History Program, IEEE History Center, 445 Hoes Lane, Piscataway, NJ 08854 USA or ieee-history@ieee.org. It should include identification of the specific passages to be quoted, anticipated use of the passages, and identification of the user.
It is recommended that this oral history be cited as follows:
Sadaoki Furui, an oral history conducted in 2021 by Mary Ann Hellrigel, IEEE History Center, Piscataway, NJ, USA.
Interview
INTERVIEWEE: Sadaoki Furui
INTERVIEWER: Mary Ann Hellrigel
DATE: 20 December 2021
PLACE: Virtual
Hellrigel:
Today is December 20th, 2021. I am Mary Ann Hellrigel, the Archivist, Institutional Historian, and Oral History Program Manager at the IEEE History Center. I am with Dr. Sadaoki Furui and we are recording his oral history virtually via WebEx from my home in New Jersey.
Hellrigel:
You’re in in Tokyo, Japan?
Furui:
Yes.
Hellrigel:
We’ve got the time difference, so technically it is December 21st in your country?
Furui:
Yes. It is 9 a.m. in Tokyo.
Hellrigel:
Yes, it is 9 a.m. in Tokyo on December 21st and it is 7 p.m. in New York City on December 20th. Thank you very much, sir. You’ve been highly recommended by the IEEE Signal Processing Society to record your oral history, and I really appreciate you taking this opportunity to work with us. I understand you’re born in 1945?
Furui:
Yes.
Hellrigel:
Now you’re an IEEE Life Fellow.
Furui:
Yes.
Hellrigel:
An IEEE Life Member and an IEEE Life Fellow, too.
Furui:
Yes.
Hellrigel:
I want to get a little bit of background information. I understand you have one younger brother who’s a theoretical physicist.
Furui:
Yes.
Hellrigel:
He is now teaching in Japan after a career in Germany?
Furui:
Yes, right.
Hellrigel:
What is his name?
Furui:
Sadakaka Furui. Mine is Sadaoki and my brother is Sadataka
Hellrigel:
Okay. He ends in an A?
Furui:
Yes.
Hellrigel:
I understand your mother and father raised you after World War II. What was your mother’s name?
Furui:
Yasuko Furui. Yes. My mom’s maiden name was Yasuko Bada, Baba. Her father had the name Masao Baba.
Hellrigel:
Your father’s name?
Furui:
Sadamasa Furui, Sadamasa. All the male’s names of my family have “sada”. They start with sada; Sadakaoki, Sadakaka and my father Sadamasa.
Hellrigel:
What your mother’s education?
Furui:
My mother studied at college. I don’t know exactly what she studied, but I heard she was studying basically the household, domestic arts, cultural and provider. Those kinds of things. I don’t know exactly what she was focused mostly on.
Hellrigel:
What kind of education did your father receive?
Furui:
He received a professional military study. Yes.
Hellrigel:
I understand that after the war, he received some additional training at a military base in Kentucky, in the United States. Did you get to travel to Kentucky with him?
Furui:
Oh, no, no, no. He was in Kentucky when I was five years old.
Hellrigel:
Oh, okay.
Furui:
He was in Kentucky from 1950 to 1951.
Hellrigel:
In the information you sent me, you said that he liked horse racing and bourbon.
Furui:
Yes.
Hellrigel:
He picked that up in Kentucky?
Furui:
Yes, yes, yes.
Hellrigel:
What did your father and mother expect for you?
Furui:
They expected me to become scientist. As I wrote in my draft, they had very sad days after the Second World War because my grandfather was executed for his responsibility for what happened in Borneo Island in the Pacific. There is much on this the Sandakan tragedy. And that is a very sad thing. Since then, my family had a very difficult time and sad days.
They wanted to raise me as kind of international, having some international collaboration scale. I was taught English. I was in elementary school and my mother sent me to a place, an English school and told me to speak in English and so.
Hellrigel:
English.
Furui:
Yes.
Hellrigel:
That was her father, so she must have been very sad.
Furui:
Yes, very sad.
Hellrigel:
Do you remember your grandfather because you were only three or four years old?
Furui:
No, no, no. My grandfather was executed when I was only one year old.
Hellrigel:
One year old.
Furui:
Yes. 1947. I was almost two years old, anyway. Yes.
Hellrigel:
After the war, how common was it that other parents would have wanted their children to be involved in international cooperation or was your family atypical?
Furui:
Maybe atypical—atypical, yes.
Hellrigel:
You learned English early on and you’re very active in what historians would call the peace movement and the reconciliation movement.
Furui:
Yes, I have been working with Malaysian and Australian people to achieve reconciliation. This is very difficult because there are many people having different kinds of ideas and opinions about how to make it possible.
Hellrigel:
This work has you travelling often to give speeches and for other activities.
After World War II, when you were growing up, Japan was rebuilding.
Furui:
Yes. Yes.
Hellrigel:
You personally had other challenges as a young boy in the period after World War II.
Furui:
Yes. It was very challenging because when I was a small kid, we had nothing to eat and we had to move to the countryside. I was born in Tokyo, but we had nothing to eat. We moved to the countryside to beg farmers to give us something to eat. We had a very difficult the time. But my mother was interested in giving me a very good education. I was very lucky and I studied very hard.
Hellrigel:
And you mentioned that you like to read that monthly magazine, The Science Magazine for children. Did you read anything else? What else did you like to do as a kid?
Furui:
Of course, I read some Japanese novels. And yes, just the general topics that I was more interested in science, so I enjoyed reading that science magazine, “Kodomo-no Kagaku” in Japanese, every month.
Hellrigel:
Your brother had a similar interest?
Furui:
Yes. Yes, he had the same, yes. I mean, we received a very big influence from my mother. Yes, my mother was really pushy to study.
Hellrigel:
Yes. Yes. She directed you.
Furui:
Yes. Yes.
Hellrigel:
And with just two children, she could keep an eye on two at the same time. If she had more, you might have been able to run away and cause trouble as little kids sometimes do.
Furui:
Yes.
Hellrigel:
You enjoyed school?
Furui:
Yes. I enjoyed school pretty much. Actually, I was studying at the elementary school and junior high that were affiliated with the college where my mother studied.
Hellrigel:
She may have even known some of your teachers.
Furui:
Yes. Yes, I think so.
Hellrigel:
Did you have any subjects you did not like?
Furui:
I was not good doing gymnastics.
Hellrigel:
Gymnastics.
Furui:
I was not good at it.
Hellrigel:
Did you play other sports?
Furui:
Yes. I was not good at sports sometime. Later, I started to play tennis, like many years later. But when I was kid, I was not good at doing any kind of sport.
Hellrigel:
You also mentioned that you play the flute. Did you start that as a kid?
Furui:
Yes. The thing is, I started practicing the flute when I was in junior high.
Hellrigel:
And were you in any clubs or activities at school?
Furui:
Yes. I was in the brass band at the junior high, and I was in the orchestra in senior high. I enjoyed it very much.
Hellrigel:
When you were a youngster your mother’s focus on education meant you did not have to think about whether you would go to college. She said, you’re going to keep going to school.
Furui:
Yes, I was supposed to study very hard and I was supposed to go to junior high, senior high school, and then university. I was educated in that way.
Hellrigel:
Did you find school highly competitive? Did your mother expect you to continue improving so you could go to increasingly better schools and then a top university?
Furui:
Yes, it’s very, very competitive in Japan. Yes, we are very competitive in getting into better and high-level schools, so I was taught how to study very hard.
Hellrigel:
Then you’re going to select the college you will attend, so how did you do that?
Furui:
I studied at the University of Tokyo. It was the highest and best university in Japan ultimately.
Hellrigel:
Did you live at school or live at home?
Furui:
Simply because I was in Tokyo I did not get into dormitory. People coming from outside of Tokyo had to choose a dormitory.
Hellrigel:
You’re at the University of Tokyo and you started there about 1963, 1964?
Furui:
Oh, 1964.
Hellrigel:
Okay, that’s what I thought. That’s a very lively time. I know that New York had the World’s Fair in 1964.
Furui:
We had the Olympic games in Tokyo in 1964.
Hellrigel:
That’s right. Did you go?
Furui:
No, no, I did not go to the games.
Hellrigel:
The whole city must have been very lively.
Furui:
Very lively that time, yes.
Hellrigel:
Maybe even the Beatles came for concert? Tokyo was doing pretty well and the economy was doing well.
Furui:
Yes, very well. We started a so-called bullet train at Tokyo at that time.
Hellrigel:
Oh, that’s right.
Furui:
We started to build the network of highways at that time. We were many years behind the United States, but we tried to catch up to these highway systems and bullet train systems.
Hellrigel:
After the war, Japan made a big effort in steelmaking, automobiles, and electronics, so it was a time for science and engineering.
Furui:
Yes, yes. Good. Yes, exactly, exactly. Yes, we were lucky.
Hellrigel:
When you’re at the University of Tokyo did the government pay or did your parents have to pay tuition? Furui: We had to pay, but it was very reasonable.
Hellrigel:
You’re at the University of Tokyo, and you’re going to study statistics?
Furui:
Yes, yes. We had to choose an expertise when we got into the fourth grade, so I chose statistics during the fourth grade of university.
Hellrigel:
You’re very good in mathematics?
Furui:
Yes. I was interested in mathematics, yes, right.
Hellrigel:
You’re going to study there, and were there many women in your class or were they all gentlemen?
Furui:
Almost all men. We had a very small number of girls, but they were very excellent, excellent.
Hellrigel:
Do you have any friends or professors from that time that were influential?
Furui:
Very small number. At the university we had a very good young professor, who was one of the pioneers of what we currently call artificial intelligence. Shun-ichi Amari was one of the pioneers artificial intelligence at the time. In 1968 or 1969, he wrote a very nice, great, nice paper. I was interested in working on kind of the very second generation or no, first to second generation of artificial intelligence. That’s very different from current technology. I enjoyed working on that and the study on that.
Hellrigel:
Do you remember his name?
Furui:
Shun-ichi Amari.
Hellrigel:
Were you able to work closely with him? Sometimes in the United States the introductory classes are so large, like 100 to 500 students.
Furui:
No. Tokyo University is not so as of course, we have very good—large classes at the first year, second year of university, but they’re going up—going up, the class is getting smaller. And when I chose expertise and the fourth year of university, the class was very small, so we studied very intensively. Yes.
Hellrigel:
At this point, you’re studying and you’re going to graduate. Were you going to go right on to work or were you going to continue your education? What’s going through your mind?
Furui:
We have to decide whether to choose working after finishing the undergrad education. I was more interested in studying, grad studying and going to graduate school more than immediately working after the undergrad. I chose a laboratory, a professor working on statistical signal processing. That is the beginning of my current career, continuing to what I have been doing until now.
Hellrigel:
At the time, was this a popular field? It’s sort of in the beginning.
Furui:
It is s the very beginning, so it was not as popular as other areas.
Hellrigel:
Right.
Furui:
Yes. Most people were working on more general topics on electronics and electrical engineering.
Hellrigel:
You are going to work on your master’s now.
Furui:
Yes.
Hellrigel:
You’re also continuing statistical work.
Furui:
Right.
Hellrigel:
What did you do for fun? What did you do outside of studying and work?
Furui:
I like traveling and I started enjoying driving. And that was the beginning of our owning car in Japan. That’s the beginning at the time. Fortunately, my parents bought a small car for me, so I started enjoying driving that car and visiting many places in Japan.
Hellrigel:
What kind of car?
Furui:
A Toyota Corolla.
Hellrigel:
Oh, Toyota has survived and the Corolla is still manufactured.
Furui:
Yes, it still is a very nice car. It is small and economical.
Hellrigel:
Right. It seems the automobile is still fun for you?
Furui:
Yes, it is still fun for me, yes.
Hellrigel:
You’re studying, but do you get to do any teaching or is it just research?
Furui:
Just research. We had no responsibility to teach when I was at grad school.
Hellrigel:
This is another two-year program for a master’s degree?
Furui:
Hellrigel:
Yes, yes, a two-year program. You have dedicated so much of your life to this, but I know at some point, Bell Labs sneaks into your world.
Furui:
Yes, yes.
Hellrigel:
This is after your master’s degree?
Furui:
Yes, after the master’s degree. While finishing my master’s degree, I had to choose to either go up to Ph.D. studies or start working at a company.
The University of Tokyo was one of the best universities in Japan, but the finances were very, very poor.
On the other hand, the NTT, Nippon Telecom Telephone was very rich, and they had their own computer to process speech waveforms.
I got interested in processing speech waveforms when I was studying for my master’s and I was interested in processing the speech waveforms by ourselves, but we had no way to do that at the university.
We had a huge computer to be shared by many people, so it could not be used for our own purposes for processing speech waveforms. I was more interested in working at NTT Lab, then continuing my studies to go higher up to Ph.D.
I started working at NTT, and as I said they had a very good facility and they had very good researchers like Bell Labs. After studying and working on research for seven years, NTT decided and offered to let me stay at Bell Labs from 1978 to 1979. It was very lucky for me. And so, I finished my Ph.D., I sent my Ph.D. thesis to the University of Tokyo before going to Bell Labs and became a doctor.
I visited Bell Labs as a Ph.D. researcher. I started working at Bell Labs with many excellent researchers and they gave me a lot of help. I enjoyed working there very much and discussing many topics, scientific topics, with many people at Bell Labs. I enjoyed driving around, and I visited New York. Bell Labs was located in Summit, New Jersey. It was very easy for me to drive or just take the bus from the Bell Labs area to New York through the Holland Tunnel. I enjoyed visiting New York many times.
Hellrigel:
Maybe the traffic wasn’t too bad on some days.
Furui:
No, not bad that time.
Hellrigel:
Did you live in Summit, New Jersey when you worked at the lab?
Furui:
I lived two or three miles from Summit, so very close to Bell Labs.
Hellrigel:
How did you like the Summit area because you left a big city like Tokyo and now you’re living in the suburbs of New Jersey?
Furui:
Yes, it’s a countryside. Yes, it is very countryside. I could see deer, but I enjoyed living there very much. We had a very good neighborhood and I enjoyed it with my family. I stayed in the house with my family, my wife and two daughters. We enjoyed living there very much.
Hellrigel:
What’s your wife’s name?
Furui:
Hikaru..
Hellrigel:
When did you get married, during grad school?
Furui:
We got married in 1971 when I just started working at NTT Labs.
Hellrigel:
I see.
Furui:
It was one year after study and I was at NTT Labs. My wife is a pianist. I enjoyed playing the flute and she helped me as a pianist. We enjoyed playing music and we got married.
Hellrigel:
You have two daughters?
Furui:
Yes.
Hellrigel:
Are they engineers or scientists?
Furui:
No, none of them is engineer or scientist.
Hellrigel:
How did they like New Jersey? They were very, very young.
Furui:
Yes. They enjoy it very much. They’re in nursery school and in kindergarten and they enjoyed making very good friends there. Yes.
Hellrigel:
You then are going to leave, but could you have stayed at the labs longer?
Furui:
No, because it was a kind of a contract and we had to return to NTT.
Hellrigel:
At this point, you’re working on speech recognition, so what did you expect to invent for the telephone company?
Furui:
Well, I started working at the basic research lab of NTT, so we were not expected to make any product. We were expected to just do basic research. We were interested in analyzing or making scientific models of how human beings are recognizing or understanding speech. We used various models, mathematics, statistics, and so forth for modeling how we human beings are handling speech. We had no hope of making a real product, a useful product, using our technology. We were just analyzing what we human beings were doing. But the technology was fortunately going forward.
I invented a very good technology at Bell Labs. It became very common to be used by almost everybody in the world to analyze speech. It was a kind of joint work of Bell Labs and NTT technology. Fortunately, that idea was very well accepted and the technology was named Delta Cepstrum by Bell Labs..
Hellrigel:
Do you own that patent or does NTT and Bell?
Furui:
Good question, because I was on leave from NTT, and I invented the technology at Bell Labs. It was very complicated about how to make the patent and we decided not to make a patent based on that technology. So, it is not patented, but instead it has been used by everybody all over the world.
Hellrigel:
That’s very open-minded.
Furui:
Yes. Very important. That’s true.
Hellrigel:
Then this research made you pretty well known ?
Furui:
Yes, yes. Fortunately, that was true. And because of that, I could make many good friends. They helped me a lot to make my good friends. Because of my Delta technology, my name became very well known to everybody working on speech through international conferences.
Hellrigel:
At this point you’re going to go back to NTT.
Furui:
Yes. NTT. Yes. Sure.
Hellrigel:
You made this big advance at Bell, so what did you expect to do next?
Furui:
I was interested in both very basic research, like human perception mechanism and human hearing mechanism. I was working on that very basic research. At the same time, I was interested in advancing my technology to make real products in the future, so I started working on two projects. One is very basic research, focusing on speech perception, and the other is speech engineering, focusing on speech recognition and speaker recognition.
In NTT headquarters, at the top, people got interested in making products, using my technology.
I worked on both basic research and engineering.
Fortunately, I had very good friends working at NTT on this both sides, basic research and engineering, and after spending twenty years at NTT Labs, I gave a one-hour talk to my bosses.
One of my bosses, a big boss on the top, asked me, you have spent already twenty years, so how many years do you need to make real useful products for speech recognition. I answered, please give me another twenty years.
Unfortunately, our speech technology that time was still very limited, and we had to deal with so many problems, spending twenty more years. Recently, we have achieved various technology using AI, and now we have speech recognition products everywhere. Actually, current AI-based speech recognition technology does not use Delta Cepstrum anymore, since similar features are automatically extracted from log-spectrum of speech waveforms in the representation leaning process of the deep-learning networks.
Hellrigel:
This may be a dopey question, but you know, that machine Alexa, that people have in their house and they ask it all sorts of questions, that’s part of your research, speech recognition.
Furui:
At the beginning, these kinds of machines used my technology, Delta Cepstrum, but probably they have changed that. They’re using completely AI-based technology these days.
Hellrigel:
It’s amazing. My sister has that and the little kids play with it. They are told, no, you’re supposed to do your homework, so don’t ask Alexa questions. Yes, that technology is amazing. I noticed that you said that was one of your dreams, to market equipment, but it was a little frustrating that it took that long. When you’re trying to develop the equipment, is there any pressure on you from NTT to put a product out so they can make money?
Furui:
Initially, there was no pressure. However, pressure was getting bigger and bigger, partly because NTT has changed. NTT has become more product oriented. Unfortunately, we were getting more pressure to make more products and not just focus on basic research. In the 1990s, I began to think about changing jobs. In 1997, I left NTT and became a professor at a university, Tokyo Institute of Technology.
Hellrigel:
How did you like that move? That’s a big move, leaving industry and going to academia.
Furui:
That’s a big move. You know, companies and universities have different cultures and different languages. I had to learn a lot: how to teach, how to do research with students, and how to conduct university management. But I basically, I enjoyed working on that. I started and I enjoyed working at Tokyo Institute of Technology because Tokyo Technology is one of the best universities in Japan, like a Japanese version of MIT. Yes, Tokyo Institute of Technology is a very good. We had very good colleagues as professor and we had very excellent students, so I enjoyed working there.
Hellrigel:
At this point, I know you have a book that was published, a very successful book, Digital Speech Processing: Synthesis, and Recognition. That must have brought you a lot of recognition. You ended up at Tokyo Institute of Technology, but were other schools looking to hire you?
Furui:
Yes. The University of Tokyo was also interested in hiring me, but I was more interested in working at Tokyo Institute of Technology. They’re more focused on science and engineering than the University of Tokyo. The University of Tokyo is a very good university covering everything. Right. So, I was more interested in working at Tokyo Institute of Technology and I chose that.
Hellrigel:
Your book is still published and now it’s in at least three languages, Japanese, English, and Chinese.
Furui:
Oh, yes, that’s true. I have never seen the Chinese version that was a kind of an informal project.
Hellrigel:
All right. You published in 1985.
Furui:
Yes.
Hellrigel:
Why did you think it was a good point to publish what became the leading textbook? Was it because there were no books out there suitable for what you wanted students to learn?
Furui:
Yes. Well, right, right. That’s true. Yes. We couldn’t find a good textbook on advanced speech processing technology. So, yes, I was interested in creating a new advanced speech textbook, covering advances and the newest technology.
Hellrigel:
Did you expect it to be around this long? It’s still in publication. You revise it for each new edition and all, but it’s like a little dynamo, it keeps going and going.
Furui:
No, actually I gave up updating that many years ago already, so I’m not updating.
Hellrigel:
At the university you have to set up a laboratory. You used to run a laboratory at NTT, but was it different at the university?
Furui:
Yes. As you can imagine, at the university you have much more freedom than at companies.
As we have discussed already, NTT has very much changed in twenty years. I wanted more freedom to enjoy and the university was much better than companies to do that. Of course, we have to spend on pay to get money, to do research and to support some people working on research.
Hellrigel:
Where do you get the funding? Is it from the Japanese government, companies?
Furui:
Mainly from the Japanese government. Yes. Right.
Hellrigel:
Like the NSF, the National Science Foundation in the United States?
Furui:
Right, right. NSF, yes. We call it JSPS in Japan.
Hellrigel:
Are you driven to get more and more grants, so you could bring in more and more students and have more projects? About how many students do you think you’ve trained over the years?
Furui:
Roughly 300, I guess.
Hellrigel:
Wow.
Furui:
Yes. Fortunately, I was internationally well-known, so we could invite and accept many students from abroad. In total, I have visited forty countries, and we have invited people, researchers, and students from thirty countries. My laboratory was very international, having student from all over the world. One of the biggest achievements that I can be very proud of is we could educate and produce very good researchers and students from my lab. They are now working at many companies like Apple, Google, Facebook and so forth. And of course, many others are working at universities, doing very good research and engineering. So that is the one of the products that I can be proud of.
Hellrigel:
Were they conducting this work in English? You’re bringing people from all over the world.
Furui:
We were speaking English, even if we were in the university in Japan. We used English every day.
Hellrigel:
Did they have to take courses, like the technical classes, in Japanese?
Furui:
That’s a good question. Yes. Most of the undergrad classes are given in Japanese, so they have difficulty if they are studying at the undergrad level. Most of my students were graduate school students, and many of the classes in graduate schools are given in English, so these courses they had no problem.
Hellrigel:
That would be a big problem for the undergrads.
Furui:
Right. Yes.
Hellrigel:
Okay. And are you still close with some of these students? That’s a big network.
Furui:
Yes. Yes. It’s fun. It’s fun. We are using Skype or Zoom these days, communicating with them. That’s fun.
Hellrigel:
And are you still active in your lab?
Furui:
No. No.
Hellrigel:
You are retired a little bit.
Furui:
Yes. I’m now advising at several universities and I’m now working mainly at two places. One is the National Institute of Informatics as chief research director, so this is a kind of advisor for the president of the research institute. The other place I am working at is the AI laboratory at the University of Tokyo. I am one of the international advisory members of the AI laboratory at the University of Tokyo.
Hellrigel:
And that keeps you busy.
Furui:
That’s fine. Yes. That’s fine.
Hellrigel:
I noticed that you talked about people who made a great impression. You gave me a short list and it included Jim Flanagan. What about him?
Furui:
He was great. He was a gentleman. He was the engineer and scientist. He was kind of the God of us. He was the department head of the Bell Labs. He influenced me and he taught me a lot. The second person was Larry Rabiner. He was also working under Jim and he was the second influential people who taught me. Aaron Rosmberg, who was very closely working with me, and Fred Juanga Taiwanese researcher. Fred also spent many years at Bell Labs, but he is now working at Georgia Tech. And Fumitada Itakura is my supervisor when I started working at NTT. Fumitada was also a great leader who taught me lot.
Hellrigel:
What did you learn from them about managing a lab?
Furui:
The most important thing I learned is how to detect what each person is good at and giving them as much freedom as possible to enhance what he or she is good at. What I have learned, I believe, is the most important thing. Yes.
Hellrigel:
You can apprise a person’s skillset and maybe help find them their best place.
Furui:
Yes. Right.
Hellrigel:
Did you have anybody say, oh, no. You don’t know what you’re doing. I’m better at this. Or did you have the skill to—you’re very softspoken. They probably said, okay, good idea.
Furui:
Honestly speaking, I was not. So sometimes I was not so successful. It was sometimes difficult to guide some people. Yes. Sometimes people are difficult. Yes.
Hellrigel:
But it seems you enjoyed it.
Furui:
Yes.
Hellrigel:
Now that you’re at the university, you not only have to manage the lab, but you also have to publish. Was that new, was this a new pressure?
Furui:
Yes. As I said, publish or perish. It is always difficult to publish good papers. I have published more than 1,000 papers. And, yes, of course, most of them are basically written with my colleagues and my students. I always help them to make it better. Fortunately, while I have been working I have been very successful publishing papers.
Hellrigel:
You’ve had a good team.
Furui:
Yes.
Hellrigel:
You know, the research lab, it’s a team.
Furui:
Yes. Yes. Right.
Hellrigel:
It is phenomenal to have 1,000 published papers.
Furui:
Yes. I was lucky. I have had very good colleagues at Tokyo Institute of Technology. I worked with a very good associate professor and assistant professors who helped me a lot.
Hellrigel:
Do you wish to single any out or name them?
Furui:
The name of the associate professor who is still working very actively is Koichi Shinoda. He has helped me a lot.
Hellrigel:
Do you still write papers?
Furui:
Me? No, no, I’m not writing any papers. One of the most recent publications was a book, titled “Universities and Society in the AI era” published in this year. That is based on what I learned, when I was working as president of TTI-Chicago (Toyota Technological Institute at Chicago) from 2013 to 2019, that was a great time for me. After retiring from the Tokyo Institute of Technology, I was elected president of TTIC. And, of course, I had to learn a lot, because I had experiences doing research at foreign countries, but I had never managed research groups in the United States. I had no idea about how to do it at universities. I had to learn a lot and, that based on what I have I learned doing present of TTIC, I decided to write a book to trigger how to improve universities and society in Japan. I have received very good responses after publishing the book. I have given some lectures based on the book at several places these days.
Hellrigel:
When you’re working for this operation based in Chicago, were you living in Chicago?
Furui:
Yes, I was living on the campus of University of Chicago. Hyde Park is the place that University of Chicago is located.
Hellrigel:
How did you like Chicago?
Furui:
I liked Chicago very much. Yes. It’s a big city and very convenient. But too flat, too windy, and too cold in the winter.
Hellrigel:
Yes. It just comes off the river and the lake. Yes.
Furui:
And tornado is a program. Unfortunately. Yes.
Hellrigel:
Did you get to do any traveling when you were there?
Furui:
Yes. I was going back and forth every month between Tokyo and Chicago, and when I was in Chicago, I have visited many places in the United States.
Hellrigel:
You got to drive?
Furui:
Yes. Drive and fly. Hellrigel: Yes. If you got out of Chicago and went to Iowa, you would have seen that state is even flatter. Yes, the Midwest.
Furui:
Yes.
Hellrigel:
When you left the Tokyo Institute of Technology, were you looking for another adventure? Or could you have stayed?
Furui:
No, I was not looking for an adventure. I left the university because of the retiring age that we have in Japan. We have to leave university at the age of 65.
Hellrigel:
It is different in the U.S.
Furui:
This is the rule that we have to follow, so I left university, but I continued working as a kind of specially appointed professor. I was given a very special class to teach Ph.D. students to make them internationalized. I had a very good, special class for teaching Ph.D. students, to make them more global. Yes. That’s very important for students, especially for Japanese students.
Hellrigel:
I noticed professionally you’re very global. I know for the IEEE Signal Processing Society and others; you’ve given many speeches and distinguished lecturers since 1993.
Furui:
Yes. Fortunately, that’s true, yes. I have had many opportunities given by my good friends. They have helped me.
Hellrigel:
During the COVID pandemic the past two years, you’ve been stuck at home. Have you been doing a lot of lectures by Zoom?
Furui:
Oh, yes, not much. But sometimes, yes. That’s easier than before.
Hellrigel:
Yes.
Furui:
Plus, the limitation what we can do by tools. We cannot avoid the time different problem. That can’t be avoided.
Hellrigel:
Yes, which some people seem to forget.
Furui:
Yes.
Hellrigel:
I was talking to someone and said wait a minute. It’s 2:00 a.m. in Australia at that point. You cannot always ask that person to be on at 2:00 a.m.
Furui:
Right.
Hellrigel:
You got to change a little bit, and especially since so much growth in IEEE is outside of the Americas now. Region 10 is growing like crazy.
Furui:
Yes. Yes.
Hellrigel:
And India with a lot, lot of students from India are joining. So that’s one of the growth areas.
Furui:
Yes, that’s true.
Hellrigel:
And that brings us to professional organizations, IEEE. Did you join IEEE as a student in the 1960s? How did that work?
Furui:
No. I started to join in 1970, I think, when I started working at NTT.
Hellrigel:
Why did you want to join IEEE?
Furui:
To publish papers, to attend conferences and to make friends.
Hellrigel:
And were your bosses active? Did they act as a mentor?
Furui:
Yes. Yes.
Hellrigel:
There are other societies in Japan. That you were a member of, because, you know, IEEE is global but were there any Japanese societies that you were active in?
Furui:
IEICE, you know, a kind of Japanese version of IEEE. And Acoustical Society of Japan, because I’m also a member of Acoustic Society of America.
Hellrigel:
Yes.
Furui:
A place for people working on acoustics and speech. And then with my good colleagues at IEEE, we started a new association what we call Asia-Pacific Signal and Information Processing Association in 2009. Twelve years ago.
Hellrigel:
The Asia-Pacific Signal and Information Processing Association?
Furui:
Yes, that’s correct. This is for encouragement. It mainly focuses on encouraging the researchers and students in Asian Pacific countries.
Hellrigel:
I guess a question would be how come you need another group? Another society? Could they do better? What did you expect?
Furui:
Probably because it’s not always for students or researchers at these developing countries to travel or publish papers. We wanted to help and still we want to help make them all active in going up to the level of - - in the next future.
Hellrigel:
You’re one of the founders. Did you hold an office? Did you ever want to be president of this society or?
Furui:
I was the first president of APSIPA. Yes.
Hellrigel:
What was that like?
Furui:
It’s fun. Simply it is fun. Of course, we are struggling on many problems because we have resources that we can cover. It’s nice to encourage young researchers and students in developing countries to grow. It’s fun.
Hellrigel:
I noticed in one of your comments you’re interested in diversity. You have outreach to the developing countries and that’s certainly promoting diversity. But you’re also reaching out to men and women.
Furui:
Yes, right. Especially Asian countries it is difficult for women to work on science and engineering. It’s difficult, but it’s also worth trying.
Hellrigel:
Right. Yes. And it’s also important like you said at the local level to have things because let’s face it, IEEE conferences are expensive. And if you can do it on the local level and people don’t have to come up with that much money, it would be a good idea. And that they have their own journal that the students can publish in?
Furui:
Yes. APSIPA has its own journal.
Hellrigel:
Yes.
Furui:
Yes. It’s an open-access journal.
Hellrigel:
Open access. Okay. That’s a good one too.
Furui:
Everybody can read it for free, yes.
Hellrigel:
Yes. And so, you’re involved with that. And now I also know that in 1993 you became an IEEE Fellow and I quote, “for contributions for speech analysis, speed recognition and speaker identification.” So that’s a big label, but do you think that justifies your career?
Furui:
Yes, I think so. The key issue was Delta feature but we have been trying not only that feature but also various technology and models to improve speech recognition and speaker recognition. Yes.
Hellrigel:
I like to ask people where were you? Do you remember what you were doing when you found out you became an IEEE Fellow?
Furui:
Sorry about meaning?
Hellrigel:
When you became an IEEE fellow, were you just, you know, eating breakfast and then you get this e-mail that says, you know, wow, you’re now a fellow. Like when people get nominated for Academy Award, they’re like, well, where were you when you got notified? Do you remember?
Furui:
Sorry, I don’t remember.
Hellrigel:
Okay.
Furui:
Maybe it was from one of my friends, my old friends.
Hellrigel:
Did you know that they nominated you or was that all done secretly?
Furui:
No. I don’t know.
Hellrigel:
What was your reaction?
Furui:
I was so happy. That was great. I did not expect for me to do that, so I was very excited.
Hellrigel:
Has it opened any new doors? What does being an IEEE Fellow mean?
Furui:
IEEE fellow means having a responsibility to teach and guide younger generations. We have to guide and teach younger researchers, yes.
Hellrigel:
You’ve been very involved in IEEE over the years. I know that now you’re an IEEE Life Fellow, which means that you successfully reached a certain age and years of membership. You’re also a member of Signal Processing Society (SPS). Were you ever a member of other societies or has it always been SPS?
Furui:
Only SPS, but I was also working with ISCA, the International Speech Communication Association. ISCA is a speech community originally founded in Europe, but we started a similar association in Asia and the United States, and they were joined together. I became one of the first presidents after joining the European and Japan and US associations together. I was not the first president, but the first after they merged. I became a president of ISCA and I have been working actively in that community.
Hellrigel:
What does working actively mean? What are you doing; presenting papers, giving speeches, and going to conferences?
Furui:
Yes. It is like IEEE, so that means attending conferences and publishing in journals. I was also one of the editors of the journal of ISCA. We have workshops and communicate with each other, so we have been working and enjoying ISCA, too. It is similar to IEEE.
Hellrigel:
In SPS you’ve received many awards. You’re one of what I call “the big fish;” you know, you’re one of the founders.
Furui:
Yes, I was lucky. Yes. I have been very lucky. Yes.
Hellrigel:
You received the James Flanagan Award which must have meant a lot since you were so close with him.
Furui:
Yes.
Hellrigel:
You also received the Signal Processing Award. What do you like about working with the Signal Processing Society? You said they’re your home community.
Furui:
Yes. We have a very good community. We’re good friends working together and open and big enough to cover many things. We have big conferences and because of that we can easily create good and new friends. So, I have been enjoying working under the Signal Processing Society.
Hellrigel:
What do you miss about not going to the conferences in person the last two years?
Furui:
Because of the COVID?
Hellrigel:
Yes, sir, because of the COVID pandemic.
Furui:
Yes. I could not attend any of the recent conferences physically. A big miss. I missed a lot of talking and hugging and drinking and eating together.
Hellrigel:
Networking.
Furui:
Yes. That’s fun and that creates new ideas even for doing research. So that sometimes we call serendipity. We can make similar opportunities also by designed Zoom kind of conferences. We can attend some special sessions, ask questions, but then it is bye-bye. Yes. We have no spare time to chat with each other.
Hellrigel:
No time to follow up, right.
Furui:
That’s very unfortunate because chatting and discussion help to create very useable products.
Hellrigel:
Right. The informal networking and chatting are lost with Zoom conferencing..
Furui:
Yes. That’s very important.
Hellrigel:
When I go to the conferences, as an historian, I also like the display room where the vendors bring in the new technology. You get to meet industry and academia and sometimes there is some type of synergy. You’ve received many awards. Where do you put them? Are they on the wall in your office? In the cabinet by behind you?
Furui:
Yes. Some are here.
Hellrigel:
Oh, cool. Excellent.
Furui:
Yes.
Hellrigel:
I see your tennis racket.
Furui:
Yes, it’s a tennis racket.
Hellrigel:
Is that your first racket?
Furui:
No. It’s a very old racket made of wood. Yes.
Hellrigel:
A wooden one, so it is older.
Furui:
Wooden one. That’s a very old, much older than I started playing.
Hellrigel:
Oh, okay. You’re involved with Signal Processing, so do you have any advice on who else I should talk to and record their oral history? Some of your friends?
Furui:
For example, Fred Juang, he’s in Atlanta at Georgia Tech. He is I think almost retiring, and he suggested me to accept the nomination for the presidency of the TTIC. Yes. He’s one of my best friends. And our current IEEE president, Ray Liu.
Hellrigel:
I was curious about your recommendations because you’ve been such a long-standing SPS member, an IEEE Fellow and now an IEEE Life Fellow. I know that you’re very active with the young people and you had pointed out that you would tell them to join IEEE for personal networks, collaboration, and working on diversity and gender issues. Do you suggest other reasons why they should join IEEE? Certainly, networking’s really important.
Furui:
Yes. We’re networking and it is very important for me. I think it’s important for anybody probably doing research and also education.
Hellrigel:
A number of times during our talk today, you also noted that it is fun working with younger IEEE members and students.
Furui:
Yes.
Hellrigel:
How do you know you’re having fun when you’re working on these big projects? What makes them fun for you?
Furui:
Of course, we cannot always be successful, so sometimes we get successful. It’s fun. Through this success we can discover what each person involved in the project is good at. What he or she is good at, as I’ve said before. We can find what he or she’s good at, and it is very important not only for the projects, but that it’s really important for himself or herself. We can help them to find what they’re good at. Yes. That is very important for developing for their future.
Hellrigel:
You’ve also had some fun with the flute and tennis, so this might be life balance for you. During your career, were you able to balance family and work?
Furui:
Yes, I think that’s important. Also, it’s very important even for doing other than teaching or doing research. It’s good and it’s important for you to have good balance. For example, I’m playing tennis with my students. That is very good for making friends and making communication with music or sports or tennis or these kind of things. We have to have good variety of ways of communication for doing the research, too. That is very important. I believe it’s very important.
Hellrigel:
When you had a young family, you said you liked to travel. Did you have fun vacations? Where did you take your family for vacation?
Furui:
Oh, that’s the biggest problem that I have had. I have been too busy to have fun as we expected, we hoped. Most of the time, for example we traveled. Of course, we enjoyed traveling at least once a year, internationally or domestically. But, unfortunately, I have been very busy and by doing many things and traveling a lot I missed some of the fun part with my family.
Hellrigel:
What have been some of your favorite IEEE conferences and other technical conferences to go to?
Furui:
Before COVID, I have attended ICASSP (International Conference on Acoustics, Speech and Signal Processing) and two or three other conferences. Since 1985, I have never missed ICASSP. That was the major conference that I attended. The major workshop or speech conference was ASRU (IEEE Workshop on Automatic Speech Recognition and Understanding).
Hellrigel:
Did you teach at these workshops?
Furui:
Oh, yes. Yes. Yes. I was general chair of ASRU twice.
Hellrigel:
Did you like that? That’s a lot of work.
Furui:
Yes. Yes. I enjoyed doing that.
Hellrigel:
Did you have any places you were able to visit for IEEE that were on your list of places you wanted to visit? When you traveled for IEEE, were there any fun places? For example, about two years ago I went to Scotland for an Applied Superconductivity Conference and then a week or so later I was back in Glasgow to attend HISTELCON and an IEEE Milestone dedication While in Glasgow I also played tourist, visiting the Kelvingrove Art Gallery and Museum, the Glasgow Transport Museum and the tall ships, the People’s Palace, and the Glasgow Cathedral and other sites. I was in Glasgow and the UK for nearly three weeks for IEEE business and a few vacation days.
Furui:
Yes. Yes. Sure. Unfortunately, I have been very busy, but every time I tried to add at least a day of extra traveling after or before the conference to enjoy something. I have tried to do that. Every time I attended conferences, I tried to stay at least one day, one extra day to visit.
Hellrigel:
Some of the IEEE conferences have companion programs. For examples when some of the IEEE boards hold meetings at the Operations Center in Piscataway I make presentations about IEEE History and provide tours of the IEEE Archives. Has your wife gone to any of these conferences?
Furui:
Yes, sometimes. Sometimes my wife joined my traveling, and she could enjoy some accompanying persons programs. For example, when we had an INTERSPEECH conference in Greece, she enjoyed visiting some beautiful Greek islands with the program.
Hellrigel:
The islands, maybe Mykonos.
Furui:
Yes, yes. Something like that. Yes. Yes.
Hellrigel:
Well, that’s fun.
Furui:
Yes, sure.
Hellrigel:
Regarding the IEEE Signal Processing Society, what do you do when you go to their conferences? Although you’re not giving papers, there was some talk about having a group discussion, a history discussion if you’d be interested in that?
Furui:
Yes. Yes. I like to study. I like to learn. I try to attend as many sessions as possible to learn something new. I just remembered the name of the IEEE president I want to ask you to conduct and oral history with, Ray Liu.
Hellrigel:
Oh, yes, K.J. Ray Liu, the 2022 IEEE president.
Furui:
Yes. He’s one of my best friends. I play tennis with him. He’s a professor at the University of Maryland. Yes.
Hellrigel:
Do you let him win?
Furui:
He’s better than I.
Hellrigel:
How about your students? When you played your students, did they let you win?
Furui:
Yes, they are sometimes kind enough to let me win.
Hellrigel:
You’re still playing today?
Furui:
Not these days. No. I’m getting old, unfortunately.
Hellrigel:
Well, it’s time for a couple of questions about the IEEE Signal Processing Society. You’ve been a member for many decades; has it changed?
Furui:
Oh, well, the big thing is growth, of course. Right. Yes. It is getting huge comparing with the old days. In the old days, the conferences were small. We could make friends easily with everybody, but it’s getting bigger and bigger. And of course, the younger population, the younger generation is growing. That’s good. But this means that I don’t know many of the people these days. That’s natural, I don’t complain, but yes, it is getting bigger.
Hellrigel:
Right. Would you have any advice for the future of SPS?
Furui:
Maybe SPS is getting too big, yes. It is covering too many areas these days. Of course, it’s not always bad. I can learn something very new by attending some of the different areas of the sessions, but the conference is getting too big. For example, I’m one of the general chairs of ICASSP next year, 2022, in Singapore. It’s getting very difficult to plan such big conferences these days. It’s not as big as the AI conferences, but these days it is very difficult to manage big conferences.
Hellrigel:
Yes, a bigger conference means you need a bigger venue and more expenses. Increased expenses, that’s a tough one. Sometimes when the new technology springs up we get a new council or a new society at IEEE. Perhaps that is one of the fears that a society has as it grows. It could break apart and go in a different direction.
Going in another direction, how important is history to you? This might seem like a weird question; however, does the history of SPS mean anything to you. For example, we are documenting SPS with these oral histories. We also have what are called firsthand histories . I can send you the link to the collection posed on Engineering and Technology History Wiki, and you can see what other people have written about their career. We’re trying to capture the history of SPS before it disappears. All too often, a society’s history can include just a list of past presidents, vice presidents, and other officers, but not really what the people did.
Furui:
I see.
Hellrigel:
What do you think is important about your history for SPS? I apologize if that seems like a dopey question.
Furui:
That’s difficult to answer. Well, I think I am one of the unusual Japanese researchers, officers working actively internationally. That’s unfortunate. In Japan, you know, most of the people are using the Japanese language for everything, and not many of them are internationalized.
Hellrigel:
Right.
Furui:
Yes. Even in education and in doing research, I always encourage professors and researchers in Japan to be more internationalized or more globalized. But it is not always easy for them to do.
Hellrigel:
Is that because of the English language or is it just the pressure to focus on Japan?
Furui:
Both. Yes.
Hellrigel:
To mark its fiftieth anniversary, SPS started an oral history collection (https://ethw.org/Oral-History:IEEE_Signal_Processing_Society_Oral_Histories). Now, to mark SPS’s seventy-fifth anniversary, we are collecting more oral histories. I will be talking to a few people suggested by SPS and some SPS members have been trained to conduct oral histories. Do you have any advice for the questions that I should ask them or you? Today, I don’t know if I’ve covered everything you wanted to cover.
Furui:
I think we have covered everything I wanted to say. Yes.
Hellrigel:
You said you’re very satisfied with your career.
Furui:
Yes, fortunately, yes. I’m a very lucky person. Yes.
Hellrigel:
Would you have done anything differently. For example, maybe you could have followed another interest say art history and become a paper maker? Do you see that you could have had a career in something else?
Furui:
Yes. As I said, at the beginning, I wanted to decide whether I could continue studying and go graduate school or start working just after the undergrad degree. When I finished my undergrad studies I was also interested in working for the Japanese government as a bureaucrat. That’s completely different than what I have done, actually. I was interested in various kinds of possibilities. I decided to continue working on research. I have been lucky and so successful, so I have no idea what would have happened if I decided to work as a bureaucrat in Japan after just finishing undergrad studies. I have no idea, but maybe I could have enjoyed something very completely different by doing that.
Hellrigel:
I could see diplomacy.
Furui:
Yes.
Hellrigel:
You were very active in the reconciliation movement and that’s a bit like diplomacy. For that type of work, you’re an ambassador really.
Furui:
Yes. I was trying, doing something like that, yes. That’s true.
Hellrigel:
You’ve made many friends. I’ve seen the photographs. You get together, but I guess COVID has delayed the face-to-face gatherings and work.
Furui:
Yes, that’s unfortunate. Yes
Hellrigel:
Will you continue this line of activity?
Furui:
Yes, if possible, as long as possible. Yes.
Hellrigel:
Are your daughters and other family members involved with this or is this you?
Furui:
My daughter is interested in helping me in the reconciliation. She wants to just part of it because she is also very busy as a Japanese bureaucrat. She’s working for the Ministry of Finance, and she’s busy. She is a very tiny bureaucrat.
Hellrigel:
Well, that’s very accomplished.
Furui:
Yes.
Hellrigel:
When you were raising your two daughters, was it clear that they were going to go to college because it seems you and your wife take education seriously?
Furui:
Yes, they studied at college and the older one is working as a bureaucrat and the younger one is working as an independent translator in English and Japanese.
Hellrigel:
They followed you a little bit, but not in careers in science and engineering
Furui:
Yes, they followed a little bit, but unfortunately, none in science or engineering. As I said, I was working at Bell Labs for one year, and my two daughters were at the nursery school and kindergarten. They enjoyed life there and quickly they got very good skills in English. Amazing
Hellrigel:
It’s the young kids that pickup language skills quickly. Yes.
Furui:
Young kids, yes. Do you know what they learned first? They said, it’s mine.
Hellrigel:
Oh, no.
Furui:
That’s the first word they learned.
Hellrigel:
Yes, in my family I joke and I call that the “myna bird.” Mine, mine, mine. It’s a kid’s skill.
Furui:
I see.
Hellrigel:
For your daughters, it was very important to learn both English and Japanese. Especially, since there are different characters or alphabets for English, Japanese, Chinese, and some other languages.
Furui:
Yes.
Hellrigel:
If you had to introduce yourself, what would you say were your three most cherished accomplishments?
Furui:
I can absolutely say, one is scientific achievement. The second achievement was, I would like to say CSJ Japanese database. You know, database is very, very important these days. What kind of databases you have changes and decides what we can do. We have been very successful in building such database, and it was almost twenty years ago that the database was built, but it is still used as the common and major speech database for Japanese speech recognition research. We are very happy. In some sense, we are very happy but also very unhappy, because it needs a huge amount of money to create such a big database. It is very difficult for current researchers to get a big budget from the Japanese government to create a new database surpassing CSJ. So, this is second. The third accomplishment that I think I can say is my good colleagues and students that I have successfully created at NTT, Tokyo Tech and TTIC. There are my treasures. I’m very happy to say they’re doing very good and are our achievements. They are now very actively working on A.I.-based speech research.
Hellrigel:
Then, one last little group of questions. You’ve been very active in IEEE. Do you have any advice for IEEE for its future? Do you have suggestions on how IEEE could improve? What is IEEE doing right?
Furui:
What we have seriously considered is how to reorganize societies. I am sure and I believe this is very difficult. Technologies are changing very rapidly and we have to adjust the structure of societies to match the progress of our research, development and engineering. For example, a society covers many things, but a society cannot cover everything related to what we are doing.
For example, A.I. is getting more and more important and we are using it for almost everything. We have related conferences and non-related conferences for A.I, but how we can cover A.I. related technology is also changing these days. We have to work on these issues. Well, we should not talk only about A.I. How we can update the structure of societies to match the progress of technology and science is very important. I believe many people are discussing these topics. It is very important for everybody to adjust the structure of societies as much as we can.
Hellrigel:
Maybe societies or the structure of societies in IEEE need to be more flexible.
Furui:
Yes.
Hellrigel:
Jose Moura said the exact same thing when I spoke to him on Monday during his oral history recording session. He said that IEEE needs to be better at staying up with the new technology going forward. Like everything, even in the field history, you got a society for this and a society for that and then a new group. New groups, fields, or sub-fields wonder if they are going to get on the conference program.
Furui:
Yes.
Hellrigel:
No one knows, you know, and so it’s like what kind of growth are you restricting?
Furui:
Yes. Jose Mora is one of my best friends. Yes.
Hellrigel:
He’s a very nice man.
Furui:
Yes, he’s a very nice man. We have been working on various opportunities, including award activities.
Hellrigel:
Yes, he is keeping busy. I spoke with him last week for his oral history because, like you, he’s SPS. He’s a Life Fellow, and then he’s a trifecta. He’s a past president. But you both have been with your organization so long and it’s interesting, almost the exact same observation. Other people that I’ve talked to also pondered on how you keep up with the change and allow that flexibility. I don’t know. Everything is society-based at the conferences. It’s a society. It’s a journal. That’s one of the challenges as you pointed out. How do you allow new growth? Maybe that’s something where IEEE’s going to have to have an ad hoc committee and say every year they do one conference on new stuff. You also pointed out the reality of the economic cost of attending these conferences, especially for those in the developing world. In addition, in the developed world the average engineer might not have funding to go to the conferences.
You’ve also been cutting edge with your society, the new society about open access. That that’s a reality, and so much of IEEE is driven by publication funding. Open access is a whole aspect of publishing in the US and internationally.
Furui:
Yes.
Hellrigel:
Yes. You’re busy and I have taken a lot of your time, but I have one last question or maybe a little bit of a discussion. Now that your career is very advanced and you’re a very senior member, what did you find the most unexpected? If you had to say to your younger self as a young member of the field and as a grad student, wow, I didn’t expect this to happen for my career. You know, wow, I have a book that’s been out for thirty years, nearly forty years. What would you say to your younger self? When you’re having a sake or a beer and you’re talking to your younger self and he’s going to say, wow.
Furui:
Probably, I can say I have tried to do my best always. I try to make as many friends as possible because what I can do myself is very limited. I wanted to create a human network as much as possible, and that helped me a lot. I’ve been so lucky I could make good friends and people. Of course, I can imagine what kind of different life that I could have tried and in completely different ways, but I cannot try to be any different anymore. So, I’m happy. I’ve been very happy and very satisfied with what I have done. And as I said already, I have always tried to do my best, and that has probably helped me make achievements.
Hellrigel:
Yes, and perseverance, I think.
Furui:
Yes.
Hellrigel:
After reading some of your other material, I noticed you just keep working. You persevered, especially, growing up in Japan at such a difficult time as a very young boy. It must be very pleasing to see the stability you created for yourself and your family.
Furui:
Yes. Yes. I would thank my family who helped me a lot and guided me.
Hellrigel:
Do you and your brother talk about technology and engineering?
Furui:
Not much because what we are doing is completely different. He’s a theoretical nuclear physicist.
Hellrigel:
Oh. Okay.
Furui:
We are doing it completely different. What he can do he’s doing is completely by theory, and sometimes he says nobody can verify that theory by experiments.
Hellrigel:
Oh, boy. Well, in so many of your work has become part of everyday life. It would be amazing if Alexander Graham Bell started the telephone with speech technology. It would be interesting, if you could have a beer with him and show him what you’ve done with speech recognition and such. That’s just a fantastic career that you’ve had.
Furui:
Yes, I’m very happy to see so many people using speech technology these days. I’m very happy to see it. I could not imagine simply what could come when I started working on speech in 1970.
Hellrigel:
Right. After I record an oral history, we still have humans make the transcription because the AI technology and letting the computer do it still has issues.
Furui:
Right.
Hellrigel:
But it’s very good for basic things.
Furui:
That’s true. Yes, that’s true.
Hellrigel:
You try to call up the credit card company, and it’s press one or say one, do this and do that.
Furui:
Yes. Yes.
Hellrigel:
My comments make it sound very simplistic, but talking to Leah Jameson, she explained there is so much work behind saying that one sentence about pressing one to do this or two to do that.
Furui:
Yes. Human beings are great. What we can do by technology still is limited, of course. We have many things to do. We human beings are doing great, and we don’t know how we are doing. As I said, the major reason why I started working on speech is to analyze how human beings recognize and understand speech. After fifty years, half a century, we can make many products, Alexa and so forth, but still, we don’t know and have no answer to my first question, how human beings recognize and understand speech. That is the reason why, as you said, what our current technology or systems can do is still limited. We have to continue working on how we human beings understand and recognize speech. We don’t know yet.
Hellrigel:
There are so many languages and different slang, colloquial terms, and metaphors, further complicate the processes.
Furui:
Yes.
Hellrigel:
For example, a phase such as “throw a monkey wrench into it.”
Hellrigel:
I’m sure the A.I. would be wondering, what are monkeys and wrenches doing in this discussion?
Furui:
Yes. Yes. Yes.
Hellrigel:
But that’s where the humans can come in and fix things.
Furui:
Yes. Yes.
Hellrigel:
One of my earliest professional experiences was working for the Thomas Edison Papers. I organized the Edison Archives and selected material for the microfilm edition (now the digital edition) of the Edison Papers. In the 1890s, Edison invented a talking doll. The doll had a little wax cylinder recording in its chest. Some people criticized the toy, claiming it sounded demonic. The reproduced voice didn’t sound like a nice little pleasant toy a little girl would play with. It sounded like a monster. The mechanism was also very delicate, so it broke easily. It was a fascinating technological thing, but it was not easy to market due to its delicate nature and the quality of the recording. However, those dolls are very collectible today and there are very few around. Edison and his team spent a lot of time selecting the best German porcelain doll with nice hair, features, and clothing. But the voices were the rough part and the internal recording and playback mechanism were the unreliable parts of the doll. This was a development from Edison’s earlier phonograph work. The talking doll was not a financial success.
Furui:
Yes.
Hellrigel:
Recently I heard a recording of Queen Victoria’s voice. It was very faint. I heard the recording via the Internet on YouTube. The recording was discovered recently and it is so interesting. It is an example of some of the earliest work on how to preserve the voice and replay it. Today, it’s digital, computerized and all of that. They really would drive themselves nuts trying to replicate the human speech back in the day with those early experiments. How can it sound?
Furui:
But now we have so-called deep fake. Yes, they can easily create fake voices, fake faces, fake bodies, and so forth, and that’s a big problem that we must discuss and we must solve.
Hellrigel:
Yes, and the ethics of it. That’s been an issue.
Furui:
Yes.
Hellrigel:
You can invent something, but if someone buys your patent, how can they use it? Maybe they are going to use it to swindle all people with these fake recordings? I’m sure you didn’t imagine that when you started fifty years ago.
Furui:
No.
Hellrigel:
You’ve had a fabulous career, and very many successes.
Furui:
Thank you.
Hellrigel:
SPS was thrilled when you could find the time to record your oral history.
Furui:
Yes. Thank you so much.
Hellrigel:
They all wish you well health-wise and everything.
Furui:
Yes. Thank you. I have to say I have made so many mistakes, so many errors in my research. I have been successful in publishing papers, but we have made many failures in doing many research that, of course, were not be published in the papers. I have been successful in some things. I have discussed only what I have been successful with, but I have made so many mistakes, so many errors, so many failures. But that’s life.
Hellrigel:
This was part of learning what you could do or cannot do?
Furui:
It is too, yes, it is. We have learned a lot from failures, yes.
Hellrigel:
If you do not mind, could you provide an example of what you consider one of your big failures, and explain why you think it’s a failure.
Furui:
One of the biggest failures, which cannot be printed or be published, is that I had a student at Tokyo Tech who committed suicide. It is my biggest mistake I could not avoid. I did not know what happened. Probably, he faced some difficulty doing his research. Unfortunately, I could not find the problem before it happened. That was a very, very sad experience that we had, and that’s the serious part. Of course, I have made many mistakes doing some research..
Hellrigel:
It would be fascinating to read a paper or have a roundtable at a conference about the research you consider a failure.
Furui:
Yes.
Hellrigel:
Well, well, some of Thomas Edison’s biggest “mistakes” or less financially successful inventions included the talking doll, but also he spent millions of dollars in iron ore mining. He founded an iron ore milling company based in the mountains of western New Jersey. He created a great magnet and it was really good technologically, but better-grade iron ore was discovered in Michigan so his multi-million-dollar empire imploded. Some refer to this as Edison’s big elephant. Technologically, it was very successful, but monetarily, not so much. Nevertheless, Edison said he had fun and would do it again because it was a great experience.
Furui:
I see.
Hellrigel:
Well, that’s very honest of you to recognize the problems. Overall, I’m sure your students are very fortunate to have you.
Furui:
I hope so.
Hellrigel:
You teach them you work in the lab with them, and you also play tennis with them and socialize with them. They must find that this makes you more accessible and approachable because you are mentoring them. For that I think they’re quite fortunate because sometimes professors don’t want to take time for their students.
Furui:
Yes.
Hellrigel:
Do you want to cover anything we did not discuss?
Furui:
No. Thank you for granting me your time.
Hellrigel:
Thank you, sir. I’ll get the recording, it will be transcribed into a Word document, I review and edit it, and then I’ll send it to you for corrections.
Furui:
Yes.
Hellrigel:
If you desire, we can schedule a second session to expand your oral history and address topics we overlooked.
Furui:
Yes, okay. I hope not.
Hellrigel:
Okay. Well, I hope not because I know I took a lot of your time, and I appreciate it, sir, because you’re one of our most esteemed members of SPS and IEEE
Furui:
Thank you.
Hellrigel:
I know SPS is just thrilled.
Furui:
Thank you so much.
Hellrigel:
Thank you, sir, and have a good day.
Furui:
Thank you so much. Have a good day. A Merry Christmas, and Happy New Year.
Hellrigel:
Thank you. Happy New Year, sir. Take care.
Furui:
Yes, thank you.
Hellrigel:
Bye-bye.
Furui:
Thank you so much. Thank you.
Hellrigel:
You’re welcome. Bye-bye.
Furui:
Bye-bye.
