Shoji Makino

Tomasz M. Rutkowski, Hiromu Mori, Yoshihiro Matsumoto et al.

A new concept and an online prototype of haptic BCI paradigm are presented. Our main goal is to develop a new, alternative and low cost paradigm, with open-source hardware and software components. We also report results obtained with the novel dry EEG electrodes based signal acquisition system by g.tec, which further improves experimental comfort. We address the following points: a novel application of the BCI; a new methodological approach used compared to earlier projects; a new benefit for potential users of a BCI; the approach working online/in real-time; development of a novel stimuli delivery hardware and software. The results with five healthy subjects and discussion of future developments conclude this submission.

Li Li, Hirokazu Kameoka, Shoji Makino

This paper proposes a new source model and training scheme to improve the accuracy and speed of the multichannel variational autoencoder (MVAE) method. The MVAE method is a recently proposed powerful multichannel source separation method. It consists of pretraining a source model represented by a conditional VAE (CVAE) and then estimating separation matrices along with other unknown parameters so that the log-likelihood is non-decreasing given an observed mixture signal. Although the MVAE method has been shown to provide high source separation performance, one drawback is the computational cost of the backpropagation steps in the separation-matrix estimation algorithm. To overcome this drawback, a method called "FastMVAE" was subsequently proposed, which uses an auxiliary classifier VAE (ACVAE) to train the source model. By using the classifier and encoder trained in this way, the optimal parameters of the source model can be inferred efficiently, albeit approximately, in each step of the algorithm. However, the generalization capability of the trained ACVAE source model was not satisfactory, which led to poor performance in situations with unseen data. To improve the generalization capability, this paper proposes a new model architecture (called the "ChimeraACVAE" model) and a training scheme based on knowledge distillation. The experimental results revealed that the proposed source model trained with the proposed loss function achieved better source separation performance with less computation time than FastMVAE. We also confirmed that our methods were able to separate 18 sources with a reasonably good accuracy.

Sota Misawa, Norihiro Takamune, Tomohiko Nakamura et al.

Rank-constrained spatial covariance matrix estimation (RCSCME) is a method for the situation that the directional target speech and the diffuse noise are mixed. In conventional RCSCME, independent low-rank matrix analysis (ILRMA) is used as the preprocessing method. We propose RCSCME using independent deeply learned matrix analysis (IDLMA), which is a supervised extension of ILRMA. In this method, IDLMA requires deep neural networks (DNNs) to separate the target speech and the noise. We use Denoiser, which is a single-channel speech enhancement DNN, in IDLMA to estimate not only the target speech but also the noise. We also propose noise self-supervised RCSCME, in which we estimate the noise-only time intervals using the output of Denoiser and design the prior distribution of the noise spatial covariance matrix for RCSCME. We confirm that the proposed methods outperform the conventional methods under several noise conditions.

Li Li, Hirokazu Kameoka, Shoji Makino

This paper proposes an alternative algorithm for multichannel variational autoencoder (MVAE), a recently proposed multichannel source separation approach. While MVAE is notable in its impressive source separation performance, the convergence-guaranteed optimization algorithm and that it allows us to estimate source-class labels simultaneously with source separation, there are still two major drawbacks, i.e., the high computational complexity and unsatisfactory source classification accuracy. To overcome these drawbacks, the proposed method employs an auxiliary classifier VAE, an information-theoretic extension of the conditional VAE, for learning the generative model of the source spectrograms. Furthermore, with the trained auxiliary classifier, we introduce a novel algorithm for the optimization that is able to not only reduce the computational time but also improve the source classification performance. We call the proposed method "fast MVAE (fMVAE)". Experimental evaluations revealed that fMVAE achieved comparative source separation performance to MVAE and about 80% source classification accuracy rate while it reduced about 93% computational time.

Hirokazu Kameoka, Li Li, Shota Inoue et al.

This paper proposes a multichannel source separation technique called the multichannel variational autoencoder (MVAE) method, which uses a conditional VAE (CVAE) to model and estimate the power spectrograms of the sources in a mixture. By training the CVAE using the spectrograms of training examples with source-class labels, we can use the trained decoder distribution as a universal generative model capable of generating spectrograms conditioned on a specified class label. By treating the latent space variables and the class label as the unknown parameters of this generative model, we can develop a convergence-guaranteed semi-blind source separation algorithm that consists of iteratively estimating the power spectrograms of the underlying sources as well as the separation matrices. In experimental evaluations, our MVAE produced better separation performance than a baseline method.

Daiki Aminaka, Shoji Makino, Tomasz M. Rutkowski

We present results of an approach to a code-modulated visual evoked potential (cVEP) based brain-computer interface (BCI) paradigm using four high-frequency flashing stimuli. To generate higher frequency stimulation compared to the state-of-the-art cVEP-based BCIs, we propose to use the light-emitting diodes (LEDs) driven from a small micro-controller board hardware generator designed by our team. The high-frequency and green-blue chromatic flashing stimuli are used in the study in order to minimize a danger of a photosensitive epilepsy (PSE). We compare the the green-blue chromatic cVEP-based BCI accuracies with the conventional white-black flicker based interface.

Kensuke Shimizu, Shoji Makino, Tomasz M. Rutkowski

The paper presents a study of an inter-stimulus interval (ISI) influence on a tactile point-pressure stimulus-based brain-computer interface's (tpBCI) classification accuracy. A novel tactile pressure generating tpBCI stimulator is also discussed, which is based on a three-by-three pins' matrix prototype. The six pin-linear patterns are presented to the user's palm during the online tpBCI experiments in an oddball style paradigm allowing for "the aha-responses" elucidation, within the event related potential (ERP). A subsequent classification accuracies' comparison is discussed based on two ISI settings in an online tpBCI application. A research hypothesis of classification accuracies' non-significant differences with various ISIs is confirmed based on the two settings of 120 ms and 300 ms, as well as with various numbers of ERP response averaging scenarios.

Chisaki Nakaizumi, Shoji Makino, Tomasz M. Rutkowski

This study provides a comprehensive test of a head-related impulse response (HRIR) cues for a spatial auditory brain-computer interface (saBCI) speller paradigm. We present a comparison with the conventional virtual sound headphone-based spatial auditory modality. We propose and optimize the three types of sound spatialization settings using a variable elevation in order to evaluate the HRIR efficacy for the saBCI. Three experienced and seven naive BCI users participated in the three experimental setups based on ten presented Japanese syllables. The obtained EEG auditory evoked potentials (AEP) resulted with encouragingly good and stable P300 responses in online BCI experiments. Our case study indicated that users could perceive elevation in the saBCI experiments generated using the HRIR measured from a general head model. The saBCI accuracy and information transfer rate (ITR) scores have been improved comparing to the classical horizontal plane-based virtual spatial sound reproduction modality, as far as the healthy users in the current pilot study are concerned.

Takumi Kodama, Shoji Makino, Tomasz M. Rutkowski

We aim at an augmentation of communication abilities of amyotrophic lateral sclerosis (ALS) patients by creating a brain-computer interface (BCI) which can control a computer or other device by using only brain activity. As a method, we use a stimulus-driven BCI based on vibration stimuli delivered via a gaming pad to the user's back. We identify P300 responses from brain activity data in response to the vibration stimuli. The user's intentions are classified according to the P300 responses recorded in the EEG. From the results of the psychophysical and online BCI experiments, we are able to classify the P300 responses very accurately, which proves the effectiveness of the proposed method.

Chisaki Nakaizumi, Toshie Matsui, Koichi Mori et al.

This study provides a comprehensive test of the head-related impulse response (HRIR) to an auditory spatial speller brain-computer interface (BCI) paradigm, including a comparison with a conventional virtual headphone-based spatial auditory modality. Five BCI-naive users participated in an experiment based on five Japanese vowels. The auditory evoked potentials obtained produced encouragingly good and stable P300-responses in online BCI experiments. Our case study indicates that the auditory HRIR spatial sound paradigm reproduced with headphones could be a viable alternative to established multi-loudspeaker surround sound BCI-speller applications.

Chisaki Nakaizumi, Koichi Mori, Toshie Matsui et al.

The aim of this study is to provide a comprehensive test of head related impulse response (HRIR) for an auditory spatial speller brain-computer interface (BCI) paradigm. The study is conducted with six users in an experimental set up based on five Japanese hiragana vowels. Auditory evoked potentials resulted with encouragingly good and stable "aha-" or P300-responses in real-world online BCI experiments. Our case study indicated that the auditory HRIR spatial sound reproduction paradigm could be a viable alternative to the established multi-loudspeaker surround sound BCI-speller applications, as far as healthy pilot study users are concerned.

Shota Kono, Daiki Aminaka, Shoji Makino et al.

The presented study explores the extent to which tactile-force stimulus delivered to a hand holding a joystick can serve as a platform for a brain computer interface (BCI). The four pressure directions are used to evoke tactile brain potential responses, thus defining a tactile-force brain computer interface (tfBCI). We present brain signal processing and classification procedures leading to successful interfacing results. Experimental results with seven subjects performing online BCI experiments provide a validation of the hand location tfBCI paradigm, while the feasibility of the concept is illuminated through remarkable information-transfer rates.

Hiromu Mori, Shoji Makino, Tomasz M. Rutkowski

The presented study explores the extent to which tactile stimuli delivered to five chest positions of a healthy user can serve as a platform for a brain computer interface (BCI) that could be used in an interactive application such as robotic vehicle operation. The five chest locations are used to evoke tactile brain potential responses, thus defining a tactile brain computer interface (tBCI). Experimental results with five subjects performing online tBCI provide a validation of the chest location tBCI paradigm, while the feasibility of the concept is illuminated through information-transfer rates. Additionally an offline classification improvement with a linear SVM classifier is presented through the case study.

Hiromu Mori, Yoshihiro Matsumoto, Victor Kryssanov et al.

The study presented explores the extent to which tactile stimuli delivered to the ten digits of a BCI-naive subject can serve as a platform for a brain computer interface (BCI) that could be used in an interactive application such as robotic vehicle operation. The ten fingertips are used to evoke somatosensory brain responses, thus defining a tactile brain computer interface (tBCI). Experimental results on subjects performing online (real-time) tBCI, using stimuli with a moderately fast inter-stimulus-interval (ISI), provide a validation of the tBCI prototype, while the feasibility of the concept is illuminated through information-transfer rates obtained through the case study.

Moonjeong Chang, Nozomu Nishikawa, Zhenyu Cai et al.

The paper presents a pilot study conducted with spatial visual, audiovisual and auditory brain-computer-interface (BCI) based speller paradigms. The psychophysical experiments are conducted with healthy subjects in order to evaluate a difficulty and a possible response accuracy variability. We also present preliminary EEG results in offline BCI mode. The obtained results validate a thesis, that spatial auditory only paradigm performs as good as the traditional visual and audiovisual speller BCI tasks.

Nozomu Nishikawa, Yoshihiro Matsumoto, Shoji Makino et al.

The paper presents results from a project aiming to create horizontally distributed surround sound sources and virtual sound images as auditory BCI (aBCI) stimuli. The purpose is to create evoked brain wave response patterns depending on attended or ignored sound directions. We propose to use a modified version of the vector based amplitude panning (VBAP) approach to achieve the goal. The so created spatial sound stimulus system for the novel oddball aBCI paradigm allows us to create a multi-command experimental environment with very encouraging results reported in this paper. We also present results showing that a modulation of the sound image depth changes also the subject responses. Finally, we also compare the proposed virtual sound approach with the traditional one based on real sound sources generated from the real loudspeaker directions. The so obtained results confirm the hypothesis of the possibility to modulate independently the brain responses to spatial types and depths of sound sources which allows for the development of the novel multi-command aBCI.

Zhenyu Cai, Shoji Makino, Takeshi Yamada et al.

The paper presents our recent results obtained with a new auditory spatial localization based BCI paradigm in which the ERP shape differences at early latencies are employed to enhance the traditional P300 responses in an oddball experimental setting. The concept relies on the recent results in auditory neuroscience showing a possibility to differentiate early anterior contralateral responses to attended spatial sources. Contemporary stimuli-driven BCI paradigms benefit mostly from the P300 ERP latencies in so called "aha-response" settings. We show the further enhancement of the classification results in spatial auditory paradigms by incorporating the N200 latencies, which differentiate the brain responses to lateral, in relation to the subject head, sound locations in the auditory space. The results reveal that those early spatial auditory ERPs boost online classification results of the BCI application. The online BCI experiments with the multi-command BCI prototype support our research hypothesis with the higher classification results and the improved information-transfer-rates.

Yoshihiro Matsumoto, Nozomu Nishikawa, Takeshi Yamada et al.

We propose a method for an improvement of auditory BCI (aBCI) paradigm based on a combination of ASSR stimuli optimization by choosing the subjects' best responses to AM-, flutter-, AM/FM and click-envelope modulated sounds. As the ASSR response features we propose pairwise phase-locking-values calculated from the EEG and next classified using binary classifier to detect attended and ignored stimuli. We also report on a possibility to use the stimuli as short as half a second, which is a step forward in ASSR based aBCI. The presented results are helpful for optimization of the aBCI stimuli for each subject.

Hiromu Mori, Yoshihiro Matsumito, Shoji Makino et al.

The paper presents results from a psychophysical study conducted to optimize vibrotactile stimuli delivered to subject finger tips in order to evoke the somatosensory responses to be utilized next in a haptic brain computer interface (hBCI) paradigm. We also present the preliminary EEG evoked responses for the chosen stimulating frequency. The obtained results confirm our hypothesis that the hBCI paradigm concept is valid and it will allow for rapid stimuli presentation in order to improve information-transfer-rate (ITR) of the BCI.