LAB LOGA page on LabLog, a review site (not the lab's official page)
Knowledge Science Lab (Komatani Lab)
Machine-translated · View original
A lab that builds systems that converse with people by voice.
AboutAbout the lab
Systems that grow smarter as they talk
The Knowledge Science Lab (Komatani Lab) studies spoken dialogue systems that talk with people, using speech recognition and natural language processing technologies.
The field has both an engineering side, writing programs to build systems, and a scientific side, asking what it means for a conversation to be mutually understood.
Because it deals with the language that humans use, it also draws on insights from linguistics and social psychology.
The lab regards building spoken dialogue systems that work in real environments as a difficult challenge, because errors in speech recognition are unavoidable.
Key terms
- Speech recognition
- Technology that takes human voice as input and converts it into text
- Spoken dialogue system
- A system that converses with people using speech recognition technology
- Multimodal dialogue system
- A system that understands dialogue using cues such as facial expressions and tone of voice, not only the content of the words
NewsLab news
A paper from the lab won the Paper Award of the Japanese Society for Artificial Intelligence for fiscal 2025
A student won the Student Encouragement Award at the 88th National Convention of the Information Processing Society of Japan
The lab developed a technology that lets a dialogue system ask questions efficiently and acquire unknown words
A student won the Student Encouragement Award at the 87th National Convention of the Information Processing Society of Japan
ResearchResearch themes
Aiming for "systems that grow smarter as they talk," the lab studies spoken dialogue systems that converse with people. It approaches this through knowledge acquisition, reading the other person's state, handling errors and breakdowns, and adaptation to users.
*Illustrative imageDialogue systems that grow smarter by talking
The knowledge of current dialogue systems is either learned from data in advance or built by hand. The lab aims for systems that grow smarter as they talk, like humans who learn new things by talking with others.
*Illustrative imageMultimodal dialogue systems that read unspoken information
To understand a dialogue, it is necessary to consider not only the content of the words but also facial expressions and tone of voice. The lab develops technology that estimates how the other person feels, using images, prosodic information, and biosignals such as EEG and skin potential.
*Illustrative imageSpoken dialogue robust to errors and breakdowns
Errors and breakdowns arise from a mismatch at one of the layers: channel, signal, intention, or dialogue. The lab aims for spoken dialogue systems that identify the cause of the mismatch, deal with it, and are robust to errors.
*Illustrative imageDialogue systems that become familiar with users
The lab develops functions that let spoken dialogue systems such as robots become "familiar with" the user through conversation. The aim is for them to hear the user's speech accurately and understand the meaning of the words the user uses without mismatch.
MembersFaculty
The members of the Knowledge Science Lab (Komatani Lab) are faculty, administrative assistants, students in the master's program and undergraduate students, and international students.As of 2026
- ProfessorKazunori Komatani
- Associate ProfessorRyu Takeda
- Assistant ProfessorKenta Yamamoto
- Assistant ProfessorBence Mark Halpern
VoicesStudent voices
Ratings and reviews reflect the experiences of individual current or former students, not the lab's official views. Whether a lab has reviews, or how many, says nothing about the lab itself.
Students who take part in an interview may receive a thank-you gift worth ¥1,000, regardless of what they say.
Student ratings (0 students)
--
--
--
--
--
--
Student voices answer questions like these
- What's great about this lab?
- What did you learn after joining?
- What's the vibe, and what is the professor like?
- Who is this lab for?
- Where do graduates go?
Your experience can help younger students choose their lab. Your name won't be shown.
Write a lab reviewAnonymous · Your draft is saved if you stopRelatedLabs similar to Knowledge Science Lab (Komatani Lab)
Integrated Intelligent Media LabSANKEN (The Institute of Scientific and Industrial Research), Research Division 1 (Information and Quantum Science)Computer vision and natural language processing as core area…
Interface Quantum Science LabSANKEN (The Institute of Scientific and Industrial Research), Research Division 1 (Information and Quantum Science)Integrating spintronic materials, nanocarbon, and flexible m…
Oiwa LabSANKEN (The Institute of Scientific and Industrial Research), Research Division 1 (Information and Quantum Science)Quantum interfaces between light and spin using single elect…
Intelligent Reasoning Lab (Shimizu Lab)SANKEN (The Institute of Scientific and Industrial Research), Research Division 1 (Information and Quantum Science)Research and development of mathematical methodologies for s…The information on this page is compiled by the LabLog team from the lab's public information and interviews. For corrections or removal requests, please contact LabLog, not the lab.
Report an errorAbout the images
The top and research theme images are illustrative and were not taken at the lab. They are cropped from Wikimedia Commons images. Cropped CC BY-SA images are provided under the same CC BY-SA license as the original (version shown on each line).
- Top image: “Arxiu audio” Jordi R. / CC BY 3.0 (cropped)
- Dialogue systems that grow smarter by talking: “Amazon Echo Dot (virtual digital assistant) with normal coffee mug as size comparison” Pittigrilli / CC BY-SA 4.0 (cropped)
- Multimodal dialogue systems that read unspoken information: “SP 25 black text” Aviad Albert / CC0 (cropped)
- Spoken dialogue robust to errors and breakdowns: “Scheme of speech recognition system” IHEARu / CC BY-SA 4.0 (cropped)
- Dialogue systems that become familiar with users: “US Navy 100531-N-7676W-075 Visitors interact with the , dexterous, social (MDS) robot Octavia at the Office of Naval Research (ONR) exhibit during Fleet Week New York 2010” U.S. Navy photo by John F. Williams / Public domain (cropped)
- Related lab (Integrated Intelligent Media Lab): “Artificial Worldviews Mapping ChatGPT” Kimfalbrecht / CC BY-SA 4.0 (cropped)
- Related lab (Interface Quantum Science Lab): “Magnetic Sensor ( )” National Institute of Standards and Technology / Public domain (cropped)
- Related lab (Oiwa Lab): “New Quantum Dot Transistor Counts Individual Photons ( )” National Institute of Standards and Technology / Public domain (cropped)
- Related lab (Intelligent Reasoning Lab (Shimizu Lab)): “Domino” Dietmar Rabich / CC BY-SA 4.0 (cropped)
About LabLog
LabLog is a lab review app for students only. Student members at each university interview students in labs and publish what they say. Student voices can only be read by students who sign up with a university email address. Operated by bestiee Inc.
Learn more about LabLog ↗Search engines and AI may reference the basic lab info and rating summaries on this page and the description of LabLog. Student voices are for student members only. Automated collection by programs (scraping, or input to or training of AI) is prohibited under Article 8 of the Terms of Service. If you need data for research or analysis, please contact us.