Ken Takaki（高木 健）
Ken Takaki is a Ph.D. student advised by the Graduate School of Engineering, the University of Tokyo. His research focuses on improving the lives of people with hearing impairments. He is exploring new approaches for wearable devices, bone conduction, and speech processing.
Ken received his Master's degree in information science and technology and a Bachelor of Engineering degree in electrical and electronic engineering from the University of Tokyo, Japan in 2021 and 2019.
He has single-sided deafness, which led him to create asEars, a glasses-type device for people with hearing loss in one ear. He is also involved in the management of Kikoiro, a community for people with hearing loss in one ear, in order to improve the lives of people with hearing loss in one ear.
Address: Room #112C1 Bldg. 2, Faculty of Eng. The Univ. of Tokyo 7-3-1 Hongo, Bunkyo-ku, Tokyo, 113-8656 JAPAN
I received my Master in information science and technology from the University of Tokyo, and entered the doctoral program.
asEars sound glasses provide enjoyable communication for people with single-sided deafness (SSD). People suffering from SSD have trouble listening to others in noisy situations, and find it difficult to enjoy their conversation. asEars collects the sound that comes from the deaf side by a pair of microphones and transmits it to the other side by a bone conduction speaker. This enables them to listen to others more clearly and enjoy communication better than ever.
work in progress
Supporting selective listening between online and offline speakers by focusing on the cocktail party effect
Online conferencing is taking the form of having conversations with people in remote locations, and sometimes with people around them. As a result, it is becoming more and more important for users to be able to concentrate on listening only to what they are interested in from the audio of online and offline speakers. In this research, we focus on the cocktail party effect and propose a speech processing method that facilitates this effect, and realize a system that facilitates selective listening to both online and offline speakers.
Project grant website (Japanese)
Coil design for wireless power transfer and communication over hinges of smart glasses
To make smart glasses foldable, we need wiring within the hinges for interconnecting the sensors and displays. This structure, however, degrades the appearance and durability of smart glasses. To solve this problem, we proposed embedding coils in the hinge’s vicinity for wireless power transfer and communication. This method can safely supply 2 W of power with 50% efficiency and can communicate at the speed of 1.7 Mbps, therefore, it can be used for eyeglass devices that present images and sounds to the user.
K. Takaki, T. Sasatani, H. Kasashima, Y. Kawahara, and T. Naemura, ”Coil design for wireless power transfer and communication over hinges of smart glasses,” in Proceedings of the 2020 International Symposium on Wearable Computers, pp.79–83, Sept. 2020.
Acoustic Length Sensor for Soft Extensible Pneumatic Actuators With a Frequency Characteristics Model
Soft robots are safe and suitable for interaction with people. On the other hand, they are easily deformed when subjected to an external force. Hence, appropriate control is necessary for their use, based on the state of the actuators represented by their length. However, conventional sensors cannot sense a broad range of length, and they have a non-linear property since they rely on the physical deformation of the sensor itself. We proposed a length sensor that can accurately measure a wide range of lengths by utilizing the acoustic characteristics of the air column inside the actuators.
K. Takaki, Y. Taguchi, S. Nishikawa, R. Niiyama and Y. Kawahara, ”Acoustic Length Sensor for Soft Extensible Pneumatic Actuators With a Frequency Characteristics Model,” in IEEE Robotics and Automation Letters, Vol. 4, No. 4, pp. 4292–4297, Oct.2019. (Young Award).
A method for estimating the bone-conducted transfer function by observing the sound radiated into the ear canal
The measurement of bone conduction hearing requires intensive operation by the user and a long measurement time, making it impractical to measure every time. In this study, we established a method to measure the bone conduction hearing every time the user wears a bone conduction device by using the bone conduction sound radiated into the ear canal. This enabled us to develop a measurement method that is quick and does not involve any manipulation by the user.