Edge AI-based Brain-Computer Interface for Real-time Applications

Henar Mike Canilang,Chigozie Uzochukwu Udeogu,James Rigor Camacho,Erick Valverde,Angela Caliwag,Wansu Lim 대한인간공학회 2021 대한인간공학회 학술대회논문집 Vol.2021 No.11

Objective: This study aims to integrate brain computer interface (BCI) to edge AI devices for real-time EEG signal processing applications. For the specific implementation in this paper, we applied edge AI device-based EEG signal processing for emotion recognition. Background: The emergence of Electroencephalogram (EEG) based applications for intelligent applications is projected to have rapid advancements in the future. The BCI system enables efficient brain signal acquisition. Current intelligent convergence of EEG based applications includes brain signal processing integrated to deep learning models. It is expected that this convergence in intelligent EEG based applications will push through to on-device local processing such as edge AI devices for portability in state-of-the-art applications. The portability and practical usage of these systems in real-world applications could lead to the development and deployment of many other advanced embedded systems for EEG-based applications. Systems that can run locally on the edge without needing to be connected to a mobile network. Edge AI devices are the leading-edge computing platforms that process data locally to overcome the current constraints of IoT application. This paves way to the integration of edge-based processing as the computing paradigm to process and acquire EEG signals. Owing to the current research advancement for both EEG and edge applications, this paper aims to propose one of the many systematic applications of deploying edge-based EEG using a brain computer interface. Method: The input for this edge-based EEG signal processing is through the BCI interfaced to the edge AI device. The edge AI device deployed with a deep learning model for specific applications locally processes the acquired signal. These acquired signals are valuable for training deep learning models to realize practical applications at the edge. The processed EEG signals enable the system response of the system such as rapid emotion recognition. Results: Varying EEG signals were acquired in each of the BCI channels. These brain signals are segmented to different brain signal clusters such as Gamma waves (30㎐ to 100㎐), Beta waves (12㎐ – 30㎐), Alpha waves (7.5㎐ – 12㎐), Theta waves (4㎐-7.5㎐) and Delta waves (0.1㎐-4㎐) which have specific brain wave description. As for EEG emotion recognition applications, these wave signals are essential for efficient and accurate emotion recognition. The alpha, beta, and gamma waves are identified to be the most discriminative frequency ranges to identify emotion. Each of the EEG signal is classified for emotion recognition and identification such as 1) valence, 2) dominance, 3) arousal and 4) liking. High and low responses from these wave signals have corresponding positive, neutral, and negative emotions based on their neural patterns at parietal and occipital sites. Other applications can use the acquired EEG signals thus maximizing the possible application of edge-based EEG signal processing. Conclusion: The local processing of the EEG signal at the edge enables the edge-based EEG system application thus enabling system response and actuation. Edge EEG also enables local and cloud co-processing whereas this maximizes the benefits of the edge computing paradigm. With this co-processing capability, it enables an adaptive and portable real-time EEG signal processing which is a constraint to conventional EEG based emotion recognition system. Application: EEG is a physiological based emotion recognition which proves to be more accurate than conventional non-physiological emotion recognition. Also, with an edge-based EEG application, it enables portability and flexibility in terms of its deployment. This application aims to be a state-of-the-art innovation to existing physiological and non-physiological emotion recognition. Furthermore, this research paper implementation aims to emphasize the vast possible applications of edge-based EEG signal processing to bridge

차량 창문 개폐 제어 사용자 인터페이스용 웨이블렛 기반 센서 신호처리 기술

민영재 대한전자공학회 2019 전자공학회논문지 Vol.56 No.12

A sensor signal processing technique for user interfaces of the vehicle window control is presented in this paper. The proposed wavelet-based sensor signal processing consists of wavelet filter banks and a user interface signal discriminator. The strength of the signal applied by the user, that is sensed according to the magnitude of the low frequency wavelet bank output signal, is transmitted to the user interface signal discriminator as a control signal. In order to improve the accuracy performance, a soft-threshold algorithm is efficiently exploited in the signal discriminator. For the practical verification and evaluation of the proposed sensor signal processing, the prototype with a piezoelectric resistance strain gauge sensor and a FPGA has been implemented. Experimental results show that the improvement of discriminant accuracy by 6.2% or more is achieved by adopting the proposed soft-threshold algorithm. With the measured accuracy over 99.3% for various user interfaces, the feasibility of the proposed sensor signal processing technology is confirmed in terms of accuracy. 본 논문에서는 차량 창문 개폐 제어 사용자 인터페이스용 센서 신호처리 기술을 제안한다. 제안하는 웨이블렛 기반 센서 신호처리 기술은 웨이블렛 필터 뱅크와 사용자 인터페이스 신호 판별부로 구성된다. 저주파 웨이블렛 뱅크 출력 신호의 크기에 따라 사용자가 인가하는 신호의 세기를 감지하고, 사용자 인터페이스 신호 판별부에 전달하여 제어 신호로 사용한다. 신호 판별부의 사용자 인가 신호에 대한 판별 정확도를 높이기 위해, 소프트 임계값을 갖는 알고리즘을 채용하였다. 제안하는 센서 신호처리 기술의 실제 검증을 위해 압전 저항 스트레인 게이지를 포함하는 센서와 FPGA의 프로토타입을 제작하였다. 실험을 통해 소프트 임계값을 이용하는 알고리즘 채용으로 최대 6.2% 이상의 판별 정확도 개선을 검증하였다. 다양한 사용자 인터페이스에 대해 99.3% 이상의 정확도를 확인하여, 본 논문에서 제안하는 센서 신호처리 기술의 정확도 측면에서의 제품 채택 가능성을 확인하였다.

실시간 초음파 영상을 위한 빔형성 및 포락선 검파의 GPU 가속 신호 처리

이원지,이명기 한국물리학회 2017 새물리 Vol.67 No.1

의료 영상 적용 분야에서 실시간 영상 재생을 위해 고속 신호 및 영상 처리는 필수 요건이다. 광음향 단층 촬영은 비침습적 방법으로 고해상도의 구조와 기능 및 분자 영상을 제공하는데, 특히, 3차원 영상 재구성, 기능 영상 및 실시간 영상 재생 등에는 빠른 신호 처리가 요구된다. 본 연구에서 우리는 실시간 영상 재생을 위해 그래픽 처리 장치(GPU)를 이용해 초음파 혹은 광음향 단층 촬영 등의 B-모드 영상을 재구성하는 고속 영상 신호 처리 방법을 제안한다. 영상 재구성에 요구되는 빔형성과 포락선 검파 과정을 대량의 GPU 코어를 이용한 병렬 계산으로 신호 처리 속도가 개선되었다. 기존의 신호 처리 방법인 중앙처리장치만(CPU)을 이용했을 경우 128 $\times$ 3200 픽셀 크기의 B-모드 영상 한 프레임의 신호 처리에 걸린 시간이 3.165 초인 반면에, GPU를 이용한 경우 평균적으로 2.778 ms로 3차원 영상 재구성 및 실시간 영상 재생 등을 위해 충분히 짧은 시간이었다. High-speed signal processing is essential for real-time displays in medical imaging applications. Photoacoustic tomography provides structural, functional, and molecular imaging with high resolution in a noninvasive way. Especially, three-dimensional image reconstruction, functional imaging, and real-time display require fast signal processing. Here, we provide a high-speed signal processing method using a graphic processing unit (GPU) to reconstruct ultrasound or photoacoustic B-mode images for real-time displays. The signal processing speed was improved by parallel processing of the beam formation and the envelop detection required for image reconstruction using a massive number of GPU cores. The time using a GPU was 2.778 ms, on average, to process a single-frame B-mode image with 128 $\times$ 3200 pixels while it was about 3.165 seconds using a central processing unit (CPU). The processing time using a GPU was short enough to reconstruct three-dimensional images for real-time displays.

적외선 배경신호 처리를 통한 OES 기반 PECVD공정 모니터링 정확도 개선

이진영,서석준,김대웅,허민,이재옥,강우석 한국반도체디스플레이기술학회 2019 반도체디스플레이기술학회지 Vol.18 No.1

Optical emission spectroscopy is used to identify chemical species and monitor the changes of process results during the plasma process. However, plasma process monitoring or fault detection by using emission signal variation monitoring is vulnerable to background signal fluctuations. IR heaters are used in semiconductor manufacturing chambers where high temperature uniformity and fast response are required. During the process, the IR lamp output fluctuates to maintain a stable process temperature. This IR signal fluctuation reacts as a background signal fluctuation to the spectrometer. In this research, we evaluate the effect of infrared background signal fluctuation on plasma process monitoring and improve the plasma process monitoring accuracy by using simple infrared background signal subtraction method. The effect of infrared background signal fluctuation on plasma process monitoring was evaluated on SiO2 PECVD process. Comparing the SiO2 film thickness and the measured emission line intensity from the by-product molecules, the effect of infrared background signal on plasma process monitoring and the necessity of background signal subtraction method were confirmed.

Performance Comparison of Different GPS L-Band Dual-Frequency Signal Processing Technologies

Kim, Hyeong-Pil,Jeong, Jin-Ho,Won, Jong-Hoon The Institute of Positioning 2018 Journal of Positioning, Navigation, and Timing Vol.7 No.1

The Global Positioning System (GPS) provides more accurate positioning estimation performance by processing L1 and L2 signals simultaneously through dual frequency signal processing technology at the L-band rather than using only L1 signal. However, if anti-spoofing (AS) mode is run at the GPS, the precision (P) code in L2 signal is encrypted to Y code (or P(Y) code). Thus, dual frequency signal processing can be done only when the effect of P(Y) code is eliminated through the L2 signal processing technology. To do this, a codeless technique or semi-codeless technique that can acquire phase measurement information of L2 signal without information about W code should be employed. In this regard, this paper implements L2 signal processing technology where two typical codeless techniques and four typical semi-codeless techniques of previous studies are applied and compares their performances to discuss the optimal technique selection according to implementation environments and constraints.

Performance Comparison of Different GPS L-Band Dual-Frequency Signal Processing Technologies

김형필,정진호,원종훈 사단법인 항법시스템학회 2018 Journal of Positioning, Navigation, and Timing Vol.7 No.1

The Global Positioning System (GPS) provides more accurate positioning estimation performance by processing L1 and L2 signals simultaneously through dual frequency signal processing technology at the L-band rather than using only L1 signal. However, if anti-spoofing (AS) mode is run at the GPS, the precision (P) code in L2 signal is encrypted to Y code (or P(Y) code). Thus, dual frequency signal processing can be done only when the effect of P(Y) code is eliminated through the L2 signal processing technology. To do this, a codeless technique or semi-codeless technique that can acquire phase measurement information of L2 signal without information about W code should be employed. In this regard, this paper implements L2 signal processing technology where two typical codeless techniques and four typical semi-codeless techniques of previous studies are applied and compares their performances to discuss the optimal technique selection according to implementation environments and constraints.

실데이터 기반 능동 소나 신호 합성 방법론

김윤수,김주호,석종원,홍정표 한국음향학회 2024 韓國音響學會誌 Vol.43 No.1

The importance of active sonar systems is emerging due to the quietness of underwater targets and the increase in ambient noise due to the increase in maritime traffic. However, the low signal-to-noise ratio of the echo signal due to multipath propagation of the signal, various clutter, ambient noise and reverberation makes it difficult to identify underwater targets using active sonar. Attempts have been made to apply data-based methods such as machine learning or deep learning to improve the performance of underwater target recognition systems, but it is difficult to collect enough data for training due to the nature of sonar datasets. Methods based on mathematical modeling have been mainly used to compensate for insufficient active sonar data. However, methodologies based on mathematical modeling have limitations in accurately simulating complex underwater phenomena. Therefore, in this paper, we propose a sonar signal synthesis method based on a deep neural network. In order to apply the neural network model to the field of sonar signal synthesis, the proposed method appropriately corrects the attention-based encoder and decoder to the sonar signal, which is the main module of the Tacotron model mainly used in the field of speech synthesis. It is possible to synthesize a signal more similar to the actual signal by training the proposed model using the dataset collected by arranging a simulated target in an actual marine environment. In order to verify the performance of the proposed method, Perceptual evaluation of audio quality test was conducted and within score difference –2.3 was shown compared to actual signal in a total of four different environments. These results prove that the active sonar signal generated by the proposed method approximates the actual signal.

Signal processing method based on energy ratio for detecting leakage of SG using EVFM

Xu Wei,Xu Ke-Jun,Yan Xiao-Xue,Yu Xin-Long,Wu Jian-Ping,Xiong Wei 한국원자력학회 2020 Nuclear Engineering and Technology Vol.52 No.8

In the sodium-cooled fast reactor, the steam generator is a heat exchange device between sodium and water, which may cause leakage, resulting in a sodium-water reaction accident, which in turn affects the safe operation of the entire nuclear reactor. To this end, the electromagnetic vortex flowmeter is used to detect leakage of the steam generator and its signal processing method is studied in this paper. The hydraulic experiment was carried out by using water instead of liquid sodium, and the sensor output signal of the electromagnetic vortex flowmeter under different gas injection volumes was collected. The bubble noise signal is reflected by the base line of the sensor output signal. According to the relationship between the proportion of the bubble noise signal in the sensor output signal and the gas injection volume, a signal processing method based on the energy ratio calculation is proposed to detect whether the water contains bubbles. The gas injection experiment of liquid sodium was conducted to verify the effectiveness of the signal processing method in the detection of bubbles in sodium, and the minimum detectable leak rate of water in the steam generator was detected to be 0.2 g/s.

Advanced signal processing for enhanced damage detection with piezoelectric wafer active sensors

Yu, Lingyu,Giurgiutiu, Victor Techno-Press 2005 Smart Structures and Systems, An International Jou Vol.1 No.2

Advanced signal processing techniques have been long introduced and widely used in structural health monitoring (SHM) and nondestructive evaluation (NDE). In our research, we applied several signal processing approaches for our embedded ultrasonic structural radar (EUSR) system to obtain improved damage detection results. The EUSR algorithm was developed to detect defects within a large area of a thin-plate specimen using a piezoelectric wafer active sensor (PWAS) array. In the EUSR, the discrete wavelet transform (DWT) was first applied for signal de-noising. Secondly, after constructing the EUSR data, the short-time Fourier transform (STFT) and continuous wavelet transform (CWT) were used for the time-frequency analysis. Then the results were compared thereafter. We eventually chose continuous wavelet transform to filter out from the original signal the component with the excitation signal's frequency. Third, cross correlation method and Hilbert transform were applied to A-scan signals to extract the time of flight (TOF) of the wave packets from the crack. Finally, the Hilbert transform was again applied to the EUSR data to extract the envelopes for final inspection result visualization. The EUSR system was implemented in LabVIEW. Several laboratory experiments have been conducted and have verified that, with the advanced signal processing approaches, the EUSR has enhanced damage detection ability.

다층구조물의 두께측정을 위한 디지털 신호처리기법

신진섭 경민대학 산학기술연구소 1999 경민대학연구논총 Vol.2 No.1

In this paper, digital signal processing technique for the thickness measurement of multilayers has been studied. The peak values of ultrasonic multiple reflected signal have been separated by power cepstrum technique, In the results of digital signal processing, Peak signals of the layers have been enhancement of visibility. The multilayers were manufactured as specimen, and the experiments for measuring its characteristics were performed, Intervals of signal which is calculated with the layer thickness compare good agreeable with result values of the digital signal processing.