Research
General Themes
-
Speech-to-speech Translation
- It has been an eternal theme for human to break through the language barrier in the communication among people who speak in a different language. The recent state of the art technology has enabled the text-to-text or speech-to-speech translation of short and simple sentences. Our laboratory will work on the speech translation of more complicated materials such as news or lectures as well as the technology to support the translation for multi-language conference attended by many persons.
-
Multi-language Communication Learning
- In case of a communication in different languages, expression varies depending on the situation and context. We conduct research on the learning of utterances and the communication support in the multi-language communication, by providing the utterances and expressions to cover such a variation.
-
Verbal, Non-verbal and Multi-modal Dialogue
- In the human-computer communication, the feasibility and the communication efficiency depend greatly on the level of userfs ability. We make a study to support the most appropriate communication by user modeling. We also work on the technology to support communication utilizing the linguistic features and intonation of the speech, the information of feeling, the facial expression and the like.
-
Individuality Modeling
- Communication requires a wide variety of individuality modeling of the users. We will go a step further and study the modeling of human voice, face, expression and dialogue modality. Furthermore, modeling the wide variety of individuality enables us to develop a communication support technology that respects and makes the most of each otherfs characteristics.
-
Concept Learning
- In order to support communication, computers need to understand not only languages but also recognize the objects found there, and, what is more, the meaning of the motion and the relevance with the words. Our research covers the study to make the computers link speech, language, image and motion, and learn the concept.
-
Multimedia Web Information Analysis
- Huge amount of various media information has been accumulated on the web. We conduct a research on the technology to analyze such a multimedia, multi-language information and extract usable information from it.
-
QoL Improvement Technology
- Improving the Quality of Life (QoL) is a very important subject for humans. Our research is extending to the technology to achieve more universal communication with supports from various aspects, aiming to realize a society where everyone is able to spend enriched life regardless of age, gender or capability.
-
Spoken Language Modeling
- Spoken language is one of the most convenient communication media.
We work on the spoken language modeling based on the statistical processing to develop the element technologies such as speech recognition, language understanding, language generation, speech conversion and speech synthesis that are sustainable under actual environment. -
Silent Speech Interface
- The normal speech utterances may cause a trouble depending on the situation, e.g. where quietness is required or the talk should not be heard by others. By using a specified sensor to record the body-conducted speech, the silent speech interface is able to transfer or input the voice via mobile phones without allowing it to be heard by others.
AHC Groups
Machine Translation (MT) |
Machine Translation Group is a study group for students who are interested in natural language processing research field, including machine translation and information retrieval.
|
|
Members: [DC2]Yusuke Oda [DC1]Philip Arthur, Akiva Miura [MC2]Takashi Sakakihara, Hiroyuki Fudaba, Makoto Morishita |
|
Spoken Dialogue (SD) |
Spoken Dialogue is a study group for students who are interested in spoken dialogue research field, including dialogue management and Q&A system.
|
|
Members: [DC3]Lasguido Nio, Masahiro Mizukami [DC2]Seitaro Shinagawa [DC1]Kyoshiro Sugiyama [MC2]Yoko Ishikawa, Jin Sasano [MC1]Seiya Kawano, Akihiro Toyoshima, Yukitoshi Murase |
|
Speech Processing (SP) |
Speech Processing is a study group for students who are interested in speech recognition(multilingual, emotion, noisy), speech understanding, speech summarization.
|
|
Members: [DC3]Kazuhiro Kobayashi, Ko Tanaka [DC2]Michael Heck [DC1]Do Quoc Truong, Takatomo Kano [MC2]Patrick Lumban Tobing, Nurul Fithria Lubis, Kinuyo Isa [MC1]Kaho Osamura, Naoki Hosomi, Takuma Mori, Tomoya Yanagita |
|
Cognitive Communication (CC) |
Cognitive Communication is a study group for students who are interested in cognitive communication reserach field, including
human brain analysis and non-verbal communication.
|
|
Members: [DC2]Hayato Maki [DC1]Hiroki Watanabe [MC2]Rui Hiraoka [MC1]Naoto Terasawa, Masahiro Honda |
|
Bigdata (BD) |
Bigdata Group is a study group for students who are interested in data engineering research field, including bigdata processing and information management.
|
|
Members: [DC2]Ikuo Keshi [MC2]Wakana Maeda [MC1]Kazuya Ikuta, Akinori Osamura, Masako Komaki, Hiroaki Tanaka, Yoshitaka Matsuda |
|
|
Publication List
-
Papers
-
Award Papers
-
Doctor Thesis
-
Master Thesis