A Non-EEG Biosignals Dataset for Assessment and Visualization of Neurological Status

Birjandtalab, Javad; Cogan, Diana; Pouyan, Maziyar Baran; Nourani, Mehrdad

doi:10.1109/sips.2016.27

Cited by 84 publications

(67 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The Stress Detection dataset [21] includes physiological non-EEG signals that are labeled with four neurological stress status. The DCNN in [2] Input is cut into frames of 64 samples.…”

Section: A Case Study 1: Stress Detectionmentioning

confidence: 99%

Minimizing Classification Energy of Binarized Neural Network Inference for Wearable Devices

Hosseini

Paneliya

Kallakuri

et al. 2019

20th International Symposium on Quality Electronic Design (ISQED)

View full text Add to dashboard Cite

In this paper, we propose a low-power hardware for efficient deployment of binarized neural networks (BNNs) that have been trained for physiological datasets. BNNs constrain weights and feature-map to 1 bit, can pack in as many 1-bit weights as the width of a memory entry provides, and can execute multiple multiply-accumulate (MAC) operations with one fused bit-wise xnor and population-count instruction over aligned packed entries. Our proposed hardware is scalable with the number of processing engines (PEs) and the memory width, both of which adjustable for the most energy efficient configuration given an application. We implement two real case studies including Physical Activity Monitoring and Stress Detection on our platform, and for each case study on the target platform, we seek the optimal PE and memory configurations. Our implementation results indicate that a configuration with a good choice of memory width and number of PEs can be optimized up to 4× and 2.5× in energy consumption respectively on Artix-7 FPGA and on 65nm CMOS ASIC implementation. We also show that, generally, wider memories make more efficient BNN processing hardware. To further reduce the energy, we introduce Pool-Skipping technique that can skip at least 25% of the operations that are accompanied by a Max-Pool layer in BNNs, leading to a total of 22% operation reduction in the Stress Detection case study. Compared to the related works using the same case studies on the same target platform and with the same classification accuracy, our hardware is respectively 4.5× and 250× more energy efficient for the Stress Detection on FPGA and Physical Activity Monitoring on ASIC, respectively.

show abstract

“…The Stress Detection dataset [21] includes physiological non-EEG signals that are labeled with four neurological stress status. The DCNN in [2] Input is cut into frames of 64 samples.…”

Section: A Case Study 1: Stress Detectionmentioning

confidence: 99%

Minimizing Classification Energy of Binarized Neural Network Inference for Wearable Devices

Hosseini

Paneliya

Kallakuri

et al. 2019

20th International Symposium on Quality Electronic Design (ISQED)

View full text Add to dashboard Cite

show abstract

“…1) Sinal Analisado: O sinal EDA foi extraído da base de dados Physionet [8] 2) Pré-processamento: O sinal de EDA foi filtrado por um passa-baixas de 0,5 Hz, para a remoção de ruídos [3]. Depois, a componente tônica foi removida pelo método dos mínimos quadrados (MMQ) em blocos contíguos.…”

Section: Configuração Experimentalunclassified

Compressão de Sinais de Atividade Eletrodérmica Através de Decomposições Atômicas Usando Dicionário de Bateman

Lima¹,

Tcheou²,

Lovisolo³

2019

Anais De XXXVII Simpósio Brasileiro De Telecomunicações E Processamento De Sinais

View full text Add to dashboard Cite

Resumo-Devido ao aumento crescente de dispositivos com sensores capazes de captar e processar sinais provenientes do corpo humano, como a atividade eletrodérmica (EDA), torna-se relevante o uso de ferramentas com capacidade de reprensentálos e comprimi-los eficientemente. Uma opçãoé o método de decomposição atômica usando o Matching Pursuit (MP) com base no dicionário de Bateman, de forma a fornecer representações esparsas. Neste trabalho, avalia-se a compressão de sinais de EDA via MP utilizando dicionário de Bateman e codificador aritmético. Palavras-Chave-Atividade Eletrodérmica, Representação Esparsa, Matching Pursuit, Codificador Aritmético, Compressão de sinais Abstract-Due to the increasing number of devices with sensors capable of capturing and processing signals from the human body, such as electrodermal activity (EDA), it becomes relevant to use tools with the ability to represent and compress them efficiently. One option is the atomic decomposition method using the Matching Pursuit (MP) based on the Bateman dictionary, in order to provide sparse representations. In this work, the compression of EDA signals via MP using the Bateman dictionary and arithmetic coder is evaluated.

show abstract

“…A. Descrição 1) Sinais Analisados: Dezoito (18) sinais de EDA extraídos da base de dados Physionet [19], [20], com aproximadamente 38 minutos de duração e taxa de amostragem de 8Hz.…”

Section: Experimentosunclassified

Análise da Decomposição de Sinais de Atividade Eletrodérmica em Múltiplos Dicionários Usando Algoritmos Vorazes

Oliveira¹,

Tcheou²,

Lovisolo³

2018

Anais De XXXVI Simpósio Brasileiro De Telecomunicações E Processamento De Sinais

View full text Add to dashboard Cite

Resumo-O grande volume de dados provenientes de sinais fisiológicos, como a atividade eletrodérmica (EDA), torna necessária a utilização de ferramentas capazes de representálos eficientemente. Uma opção são métodos de decomposição adaptativa como o Matching Pursuit (MP) e o Othogonal Matching Pursuit (OMP) aplicados com dicionários apropriados, de forma a fornecer representações esparsas. Neste trabalho, avalia-se a decomposição de sinais de EDA via MP e OMP utilizando múltiplos dicionários simultâneos/concorrentes (Bateman, Gabor e Triângulos). Os resultados indicam que utilizar múltiplos dicionários pode ser vantajoso e que, dentre os utilizados, o dicionário de Bateman é o mais relevante em termos de extração da energia de sinais de EDA. Palavras-Chave-Atividade Eletrodérmica, Representação Esparsa, Matching Pursuit, Orthogonal Matching PursuitAbstract-The large volume of data in physiological signals, like electrodermal activity (EDA), requires tools capable of representing such signals efficiently. Adaptive decomposition methods, such as Matching Pursuit and Othogonal Matching Pursuit may provide such representations if employed with the appropriate dictionary, guaranteing compact and meaningful representations. In this work, we evaluate the decomposition of EDA signals using multiple simultanoeus/concurrent dictionaries (Bateman, Gabor and Triangles) with both algorithms. Simulation results show that using multiple dictionaries may be advantageous, and that the Bateman dictionary is the most relevant one in terms of energy extraction among them for EDA signals.

show abstract

A Non-EEG Biosignals Dataset for Assessment and Visualization of Neurological Status

Cited by 84 publications

References 6 publications

Minimizing Classification Energy of Binarized Neural Network Inference for Wearable Devices

Minimizing Classification Energy of Binarized Neural Network Inference for Wearable Devices

Compressão de Sinais de Atividade Eletrodérmica Através de Decomposições Atômicas Usando Dicionário de Bateman

Análise da Decomposição de Sinais de Atividade Eletrodérmica em Múltiplos Dicionários Usando Algoritmos Vorazes

Contact Info

Product

Resources

About