Pub Date : 2018-10-01DOI: 10.1109/BIOCAS.2018.8584762
Bilal Majeed, Hafiz Talha Iqbal, Uzair Khan, Muhammad Awais Bin Altaf
Thermogram (infrared) based non-contact noninvasive screening device for breast cancer is presented. To enable a home-based patient-comfort portable breast cancer screening device, a FLIR thermal imaging camera, with the best of four (BoF) features based on the proposed novel methodology and the support vector machine (SVM) learning classifier is exploited. The 4-dimensjon Feature Vector (FV) is computed using the segmented image, grey Level co-occurrence matrix (GLCM) and run-length matrix (RLM) calculation. To ensure hardware optimization, the proposed multiplexed GLCM, RLM and SVM implementation realizes an area reduction of 30% compared to the conventional with minimal overhead in the system speed requirement. A Linear SVM is utilized to decide between malignant and benign based on the FV. The system is implemented on FPGA and experimentally verified using the patients from the Mastological Research database. The proposed breast cancer screening processor targets a portable home environment and achieves the sensitivity and specificity of 79.06% and 88.57%, respectively.
{"title":"A Portable Thermogram based Non-contact Non-invasive Early Breast-Cancer Screening Device","authors":"Bilal Majeed, Hafiz Talha Iqbal, Uzair Khan, Muhammad Awais Bin Altaf","doi":"10.1109/BIOCAS.2018.8584762","DOIUrl":"https://doi.org/10.1109/BIOCAS.2018.8584762","url":null,"abstract":"Thermogram (infrared) based non-contact noninvasive screening device for breast cancer is presented. To enable a home-based patient-comfort portable breast cancer screening device, a FLIR thermal imaging camera, with the best of four (BoF) features based on the proposed novel methodology and the support vector machine (SVM) learning classifier is exploited. The 4-dimensjon Feature Vector (FV) is computed using the segmented image, grey Level co-occurrence matrix (GLCM) and run-length matrix (RLM) calculation. To ensure hardware optimization, the proposed multiplexed GLCM, RLM and SVM implementation realizes an area reduction of 30% compared to the conventional with minimal overhead in the system speed requirement. A Linear SVM is utilized to decide between malignant and benign based on the FV. The system is implemented on FPGA and experimentally verified using the patients from the Mastological Research database. The proposed breast cancer screening processor targets a portable home environment and achieves the sensitivity and specificity of 79.06% and 88.57%, respectively.","PeriodicalId":259162,"journal":{"name":"2018 IEEE Biomedical Circuits and Systems Conference (BioCAS)","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133353874","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-10-01DOI: 10.1109/BIOCAS.2018.8584703
Nicholas H. Vitale, M. Azin, P. Mohseni
This live demonstration showcases a fully interactive, Bluetooth low energy (BLE)-enabled, wireless interface for bidirectional communications with a neural microsystem. A standalone user base station (UBS) employs custom software to wirelessly communicate with the prototype microsystem via BLE. Users can seamlessly interact with the UBS to program and monitor various features and functions of the neural microsystem wirelessly over meter-range distances.
{"title":"Live Demonstration: A Bluetooth Low Energy (BLE)-enabled Wireless Link for Bidirectional Communications with a Neural Microsystem","authors":"Nicholas H. Vitale, M. Azin, P. Mohseni","doi":"10.1109/BIOCAS.2018.8584703","DOIUrl":"https://doi.org/10.1109/BIOCAS.2018.8584703","url":null,"abstract":"This live demonstration showcases a fully interactive, Bluetooth low energy (BLE)-enabled, wireless interface for bidirectional communications with a neural microsystem. A standalone user base station (UBS) employs custom software to wirelessly communicate with the prototype microsystem via BLE. Users can seamlessly interact with the UBS to program and monitor various features and functions of the neural microsystem wirelessly over meter-range distances.","PeriodicalId":259162,"journal":{"name":"2018 IEEE Biomedical Circuits and Systems Conference (BioCAS)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114350647","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-10-01DOI: 10.1109/BIOCAS.2018.8584823
A. T. Zavareh, S. Hoyos
The ever-increasing demand for acquiring images at a faster rate and consequently of deeper layers of tissue in swept-source optical coherence tomography (SS-OCT), has motivated researchers to investigate techniques capable of performing fast acquisition without compromising sensitivity. Non-linear spectral sweeps, phase instability, and increased noise levels of swept lasers can cause the image quality to deteriorate, especially if a lower quality light source is used. This work leverages the unscented Kalman filter aiming to alleviate these shortcomings. Using this filter, both the non-idealities of the non-linear sweep and the increased noise levels are accounted for. Simulations show promising performance results in terms of extracting the non-linear spectral sweep as a function of time. The UKF also shows better metrics compared to other versions of the Kalman filter when it comes to tracking non-linearities and hardware implementation complexity giving the method significant advantages in real-time applications of hand-held, portable optical coherence tomography devices.
{"title":"The Spectral Calibration of Swept-Source Optical Coherence Tomography Systems Using Unscented Kalman Filter","authors":"A. T. Zavareh, S. Hoyos","doi":"10.1109/BIOCAS.2018.8584823","DOIUrl":"https://doi.org/10.1109/BIOCAS.2018.8584823","url":null,"abstract":"The ever-increasing demand for acquiring images at a faster rate and consequently of deeper layers of tissue in swept-source optical coherence tomography (SS-OCT), has motivated researchers to investigate techniques capable of performing fast acquisition without compromising sensitivity. Non-linear spectral sweeps, phase instability, and increased noise levels of swept lasers can cause the image quality to deteriorate, especially if a lower quality light source is used. This work leverages the unscented Kalman filter aiming to alleviate these shortcomings. Using this filter, both the non-idealities of the non-linear sweep and the increased noise levels are accounted for. Simulations show promising performance results in terms of extracting the non-linear spectral sweep as a function of time. The UKF also shows better metrics compared to other versions of the Kalman filter when it comes to tracking non-linearities and hardware implementation complexity giving the method significant advantages in real-time applications of hand-held, portable optical coherence tomography devices.","PeriodicalId":259162,"journal":{"name":"2018 IEEE Biomedical Circuits and Systems Conference (BioCAS)","volume":"101 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114517353","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Recent advances in neuromorphic computing system have shown resistive random-access memory (ReRAM) can be used to efficiently implement compact parallel computing arrays, which are inherently suitable for neural networks that require large amounts of matrix-vector multiplications (MVMs). In this work, we proposed a neuromorphic computing system based on ReRAM synaptic array to implement bitwise neural networks. The system contains a ReRAM synaptic array for parallel computation of bitwise MVMs, and a field-programmable gate array for data buffering and processing. To deploy the network on the system, a customized training scheme was required to adapt the trained network to the characteristic of ReRAM synaptic array with bitwise weights and inputs. We also managed the resolution of partial sum to reduce the bit width requirement of sense amplifier, thereby reducing power consumption. The measurement results show that the ReRAM synaptic array consumed only 0.27mW at 1V supply by using 1-bit sense amplifier while the system still maintained 97.52% accuracy on MNIST dataset.
{"title":"A Neuromorphic Computing System for Bitwise Neural Networks Based on ReRAM Synaptic Array","authors":"Pin-Yi Li, Cheng-Han Yang, Wei-Hao Chen, Jian-Hao Huang, Wei-Chen Wei, Je-Syu Liu, Wei-Yu Lin, Tzu-Hsiang Hsu, C. Hsieh, Ren-Shuo Liu, Meng-Fan Chang, K. Tang","doi":"10.1109/BIOCAS.2018.8584810","DOIUrl":"https://doi.org/10.1109/BIOCAS.2018.8584810","url":null,"abstract":"Recent advances in neuromorphic computing system have shown resistive random-access memory (ReRAM) can be used to efficiently implement compact parallel computing arrays, which are inherently suitable for neural networks that require large amounts of matrix-vector multiplications (MVMs). In this work, we proposed a neuromorphic computing system based on ReRAM synaptic array to implement bitwise neural networks. The system contains a ReRAM synaptic array for parallel computation of bitwise MVMs, and a field-programmable gate array for data buffering and processing. To deploy the network on the system, a customized training scheme was required to adapt the trained network to the characteristic of ReRAM synaptic array with bitwise weights and inputs. We also managed the resolution of partial sum to reduce the bit width requirement of sense amplifier, thereby reducing power consumption. The measurement results show that the ReRAM synaptic array consumed only 0.27mW at 1V supply by using 1-bit sense amplifier while the system still maintained 97.52% accuracy on MNIST dataset.","PeriodicalId":259162,"journal":{"name":"2018 IEEE Biomedical Circuits and Systems Conference (BioCAS)","volume":"138 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116290991","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-10-01DOI: 10.1109/BIOCAS.2018.8584801
Rajat Emanuel Singh, K. Iqbal, G. White
The organization of encoded motor modules or motor primitives in the central nervous system and their combination leads to different aspects of natural motor behavior. It is believed that neural stimulation of these coded sections activates specific groups of muscles to achieve a behavioral goal. We use the muscle synergy (MS) hypothesis to compare activation patterns during overground walking and slackline walking for a small group of highly proficient slackliners and beginners. Synchronous MS were extracted using factor analysis (FA) for rhythmic and arrhythmic repertoire of movement. The results revealed no significant difference between slackliners and non-slackliners as the extracted synergies were dependent on the variability of the task. Besides, the shared dimensional space revealed the task-specific higher loading of the quadriceps muscles for walking with such postural constraints.
{"title":"Muscle Synergy Adaptation During A Complex Postural Stabilization Task*","authors":"Rajat Emanuel Singh, K. Iqbal, G. White","doi":"10.1109/BIOCAS.2018.8584801","DOIUrl":"https://doi.org/10.1109/BIOCAS.2018.8584801","url":null,"abstract":"The organization of encoded motor modules or motor primitives in the central nervous system and their combination leads to different aspects of natural motor behavior. It is believed that neural stimulation of these coded sections activates specific groups of muscles to achieve a behavioral goal. We use the muscle synergy (MS) hypothesis to compare activation patterns during overground walking and slackline walking for a small group of highly proficient slackliners and beginners. Synchronous MS were extracted using factor analysis (FA) for rhythmic and arrhythmic repertoire of movement. The results revealed no significant difference between slackliners and non-slackliners as the extracted synergies were dependent on the variability of the task. Besides, the shared dimensional space revealed the task-specific higher loading of the quadriceps muscles for walking with such postural constraints.","PeriodicalId":259162,"journal":{"name":"2018 IEEE Biomedical Circuits and Systems Conference (BioCAS)","volume":"20 3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116702085","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-10-01DOI: 10.1109/BIOCAS.2018.8584841
K. Hayashi, S. Arata, Ge Xu, S. Murakami, C. D. Bui, A. Kobayashi, K. Niitsu
In order to care and prevent diabetes, continuous glucose monitoring (CGM) is considerably important [1]. However, existing needle-type glucose monitoring systems [2] are painful, thus unsuitable for preventing applications. In order to address this issue, a needle-less and pain-free CGM contact lens has been developed [3]. Since tear glucose level has high correlation with blood glucose level, it can be expected to be future de facto standard CGM. However, the existing CGM contact lens is based on RFID technology associated with the dedicated smart glasses, which degrade the patients' comfort and cost too much.
{"title":"Live Demonstration: 385 × 385 μm2 0.165V 270pW Fully-Integrated Supply-Modulated OOK Tx in 65nm CMOS for Glasses-Free, Self-Powered, and Fuel-Cell-Embedded Continuous Glucose Monitoring Contact Lens","authors":"K. Hayashi, S. Arata, Ge Xu, S. Murakami, C. D. Bui, A. Kobayashi, K. Niitsu","doi":"10.1109/BIOCAS.2018.8584841","DOIUrl":"https://doi.org/10.1109/BIOCAS.2018.8584841","url":null,"abstract":"In order to care and prevent diabetes, continuous glucose monitoring (CGM) is considerably important [1]. However, existing needle-type glucose monitoring systems [2] are painful, thus unsuitable for preventing applications. In order to address this issue, a needle-less and pain-free CGM contact lens has been developed [3]. Since tear glucose level has high correlation with blood glucose level, it can be expected to be future de facto standard CGM. However, the existing CGM contact lens is based on RFID technology associated with the dedicated smart glasses, which degrade the patients' comfort and cost too much.","PeriodicalId":259162,"journal":{"name":"2018 IEEE Biomedical Circuits and Systems Conference (BioCAS)","volume":"37 2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128231019","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-10-01DOI: 10.1109/BIOCAS.2018.8584771
Ahmad Rezvanitabar, Gwangrok Jung, F. Degertekin, Maysam Ghovanloo
This work presents an integrated solution for lowering the power consumption in interface circuits for bridge sensors, which consume power because of their low resistances particularly in implantable microsystems, such as intracranial pressure (ICP) monitoring application. The proposed direct bridge-to-digital interface uses a pseudo-pseudo differential (PPD) structure, in which the converter not only provides key benefits of traditional fully-differential interface circuits but also can reduce their complexity with single-ended architecture. It occupies 0.0667 mm2in 0.35-μm standard CMOS technology, where the interface provides 9.13 effective number of bits (ENOB), while cutting the power consumption of a 3 kΩWheatstone bridge down to 363 µW at 1.8 V supply, and sampling rate of 3.72 kHz in post-layout simulations.
{"title":"Toward an Energy-Efficient Bridge-to-Digital Intracranial Pressure Sensing Interface","authors":"Ahmad Rezvanitabar, Gwangrok Jung, F. Degertekin, Maysam Ghovanloo","doi":"10.1109/BIOCAS.2018.8584771","DOIUrl":"https://doi.org/10.1109/BIOCAS.2018.8584771","url":null,"abstract":"This work presents an integrated solution for lowering the power consumption in interface circuits for bridge sensors, which consume power because of their low resistances particularly in implantable microsystems, such as intracranial pressure (ICP) monitoring application. The proposed direct bridge-to-digital interface uses a pseudo-pseudo differential (PPD) structure, in which the converter not only provides key benefits of traditional fully-differential interface circuits but also can reduce their complexity with single-ended architecture. It occupies 0.0667 mm2in 0.35-μm standard CMOS technology, where the interface provides 9.13 effective number of bits (ENOB), while cutting the power consumption of a 3 kΩWheatstone bridge down to 363 µW at 1.8 V supply, and sampling rate of 3.72 kHz in post-layout simulations.","PeriodicalId":259162,"journal":{"name":"2018 IEEE Biomedical Circuits and Systems Conference (BioCAS)","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134337906","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-10-01DOI: 10.1109/BIOCAS.2018.8584766
Sylmarie Dávila-Montero, E. Ashoori, A. Mason
This paper demonstrates the use of a data decimation method, called phase-preserving quantization (PPQ), for early seizure prediction. PPQ consists of a) amplifying and filtering the neural signals around the frequency band of interest, and b) compressing the filtered signal using a 1-bit quantizer with a 0-V single-threshold decision. The ability of PPQ to retain phase information and predict seizure events while compressing the signal resolution to a single bit is demonstrated using electroencephalography (EEG) recordings from the Children's Hospital Boston-MIT (CHB-MIT) EEG database. Results show 97% accuracy when calculating synchrony values using PPQ, which is an improvement of 7% when compared to previously published results. The presented improved method enables the early detection of seizure events, resulting in a decrease in phase synchrony computation time while allowing an increase in the number of recording channels that can be screened when using EEG.
{"title":"Early Detection of Epileptic Activity on EEG Signals using Phase-Preserving Quantization Method","authors":"Sylmarie Dávila-Montero, E. Ashoori, A. Mason","doi":"10.1109/BIOCAS.2018.8584766","DOIUrl":"https://doi.org/10.1109/BIOCAS.2018.8584766","url":null,"abstract":"This paper demonstrates the use of a data decimation method, called phase-preserving quantization (PPQ), for early seizure prediction. PPQ consists of a) amplifying and filtering the neural signals around the frequency band of interest, and b) compressing the filtered signal using a 1-bit quantizer with a 0-V single-threshold decision. The ability of PPQ to retain phase information and predict seizure events while compressing the signal resolution to a single bit is demonstrated using electroencephalography (EEG) recordings from the Children's Hospital Boston-MIT (CHB-MIT) EEG database. Results show 97% accuracy when calculating synchrony values using PPQ, which is an improvement of 7% when compared to previously published results. The presented improved method enables the early detection of seizure events, resulting in a decrease in phase synchrony computation time while allowing an increase in the number of recording channels that can be screened when using EEG.","PeriodicalId":259162,"journal":{"name":"2018 IEEE Biomedical Circuits and Systems Conference (BioCAS)","volume":"68 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133576021","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-10-01DOI: 10.1109/BIOCAS.2018.8584671
Callen Votzke, U. Daalkhaijav, Y. Mengüç, M. Johnston
Stretchable electronic circuits and systems will be critical for future wearable devices and smart textiles, where existing rigid and flexible fabrication approaches severely limit conformal deformation. This is especially true for wearable sensors and actuators, critical for emerging physical human-machine interfaces and stretchable electrical interconnects. In this work, we present a 3D-printed, highly-stretchable strain sensor that uses a modified liquid metal paste to provide high-strain conductors. This approach provides near-zero hysteresis compared with nanotube-based inks, and improved conductivity over carbon- and metal-based inks, both critical for integration of soft sensors with stretchable measurement circuitry. We present an approach for fabrication of the wearable sensors and demonstrate stable conductivity of the liquid metal paste with near-zero hysteresis over 375 cycles at 200% strain. The device is demonstrated for measurement of elbow flexion angle, providing proof-of-concept of the approach for biomechanical sensor applications and wearable human-machine interfaces.
{"title":"Highly-Stretchable Biomechanical Strain Sensor using Printed Liquid Metal Paste","authors":"Callen Votzke, U. Daalkhaijav, Y. Mengüç, M. Johnston","doi":"10.1109/BIOCAS.2018.8584671","DOIUrl":"https://doi.org/10.1109/BIOCAS.2018.8584671","url":null,"abstract":"Stretchable electronic circuits and systems will be critical for future wearable devices and smart textiles, where existing rigid and flexible fabrication approaches severely limit conformal deformation. This is especially true for wearable sensors and actuators, critical for emerging physical human-machine interfaces and stretchable electrical interconnects. In this work, we present a 3D-printed, highly-stretchable strain sensor that uses a modified liquid metal paste to provide high-strain conductors. This approach provides near-zero hysteresis compared with nanotube-based inks, and improved conductivity over carbon- and metal-based inks, both critical for integration of soft sensors with stretchable measurement circuitry. We present an approach for fabrication of the wearable sensors and demonstrate stable conductivity of the liquid metal paste with near-zero hysteresis over 375 cycles at 200% strain. The device is demonstrated for measurement of elbow flexion angle, providing proof-of-concept of the approach for biomechanical sensor applications and wearable human-machine interfaces.","PeriodicalId":259162,"journal":{"name":"2018 IEEE Biomedical Circuits and Systems Conference (BioCAS)","volume":"253 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132826357","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-10-01DOI: 10.1109/BIOCAS.2018.8584746
Jaehyun Park, Ganapati Bhat, C. S. Geyik, Ümit Y. Ogras, H. Lee
Small form factor and low-cost wearable devices enable a variety of applications including gesture recognition, health monitoring, and activity tracking. Energy harvesting and optimal energy management are critical for the adoption of these devices, since they are severely constrained by battery capacity. This paper considers optimal gesture recognition using self-powered devices. We propose an approach to maximize the number of gestures that can be recognized under energy budget and accuracy constraints. We construct a computationally efficient optimization algorithm with the help of analytical models derived using the energy consumption breakdown of a wearable device. Our empirical evaluations demonstrate up to 2.4 x increase in the number of recognized gestures compared to a manually optimized solution.
{"title":"Energy-Optimal Gesture Recognition using Self-Powered Wearable Devices","authors":"Jaehyun Park, Ganapati Bhat, C. S. Geyik, Ümit Y. Ogras, H. Lee","doi":"10.1109/BIOCAS.2018.8584746","DOIUrl":"https://doi.org/10.1109/BIOCAS.2018.8584746","url":null,"abstract":"Small form factor and low-cost wearable devices enable a variety of applications including gesture recognition, health monitoring, and activity tracking. Energy harvesting and optimal energy management are critical for the adoption of these devices, since they are severely constrained by battery capacity. This paper considers optimal gesture recognition using self-powered devices. We propose an approach to maximize the number of gestures that can be recognized under energy budget and accuracy constraints. We construct a computationally efficient optimization algorithm with the help of analytical models derived using the energy consumption breakdown of a wearable device. Our empirical evaluations demonstrate up to 2.4 x increase in the number of recognized gestures compared to a manually optimized solution.","PeriodicalId":259162,"journal":{"name":"2018 IEEE Biomedical Circuits and Systems Conference (BioCAS)","volume":"20 2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130532459","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：