▼お仕事のご依頼

執筆、取材、講演のお問い合わせはこちらからお願いします。

▼広告のご依頼

MM-Labo.com内への広告のご希望はこちらからお願いします｡

当研究所の研究テーマは、生命と機械の違いとはなんであるかを理解し、生命の持つさまざまな機能を機械で実現するための技術要素を開発することにあります。

生命は現在の機械では到底及ばないようなさまざまな機能を持っています。それらの機能を一部分でも機械で実現することで、生命についてのより深い知見が得られるだけでなく、私たちの生活の中でもさまざまな役に立つ機械を作ることができるはずです。
われわれは工学的な見地から生命の機能を機械で実現することを目指し、また、その開発を通して、生命と機械の違いがどこにあるのかを理解したいと考えています。

音声処理 (Acoustical signal processing)

音源分離、音源定位に関する研究です。従来のマイクロフォンアレイではマイクロホンを異なる位置に置き、その位相差を用いることで音源分離や音源定位を実現しますが、位相差を使っているために小型化が困難であるという問題があります。提案手法では、同じ位置に複数のマイクを置き，かわりにマイクロホンの指向特性の違いを利用することで音源分離や音源定位を実現しています。マイクロホンを同位置に置くため、原理的にシステムの大きさを小さくでき，ロボットに実装することが容易であるなどの利点があります．

This research is about miniaturizing microphone array utilizing aggregated microphones. The conventional microphone array realizes the delay-sum type microphone array, and DOA estimation such as high resolution algorithms, by utilizing the phase difference of each microphone. However, it is difficult to miniaturize the microphone array due to utilization of the phase difference of each microphone.
Correspondingly, we proposed another type of microphone array that does not utilize the phase difference of each microphone but utilizes the directivities of microphones located at the same place, namely aggregated microphones.
In this method, all microphones are located in the same position, and directional microphones are arranged to differentiate the amplitude gains. Hence, the system can easily be miniaturized.
This feature is useful when applied to small robots, conference systems, and so on.

This research is about extracting the features from an image. In vector analysis, it is important to classify three flow primitives as translation, rotation and divergence. These three primitives can be detected utilizing line integral and surface integral according to the knowledge of vector analysis. This method is for extracting these three primitives utilizing edges in an image based on vector analysis, namely edge field analysis(EFA). The edge has the information of inclination. However, the edge has no information of the direction unlike vector. Hence, line integral and surface integral can not be directly applied to detect these three primitives. Applying EFA to a image, we can extract various features from a image such as the center of the storm, the center of the flower and the center of the finter print.
In general, we extract the velocity of the objects utilizing muitiple discrete images. However, applying EFA to a blurred image, we can extract quasi-motion from a blurred image.

This research is about self-reproducing hardware. Self assembly plays an important role in a variety of research fields such as robotics, chemicals, physics and so on. Utilizing self assembly, we can simplify the process of assembling the hardware. However, due to the limitation of self assembly, it is difficult for us to create the reproduction of the assembled hardware. If the advanced method of self assembly, namely self-reproduction can be realized, we will be able to obtain the same kinds of hardwares as the objective hardwares from the simple elements automatically.
We consider the study on self-assembly and self-reproduction is noteworthy not only for simplifying the process of assembling the hardware but for recognizing the life. Through this study, we would like to consider the difference between the life and machinery.