Alexander H. Nelson

Dr. Alexander H. Nelson


Research Focus & Metrics

Academic Career Length

14Years (2013-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 January 2022 in Journal of Infrared, Millimeter, and Terahertz Waves
Verified
Journal Article
Closed access
PDF file
HLHaoyan Liu
Haoyan Liu
NVNagma Vohra
Nagma Vohra
KBKeith Bailey
Keith Bailey
MEMagda El-Shenawee
Magda El-Shenawee
ANAlexander H. Nelson
Alexander H. Nelson

26 October 2015 in IEEE Transactions on Multi-Scale Computing Systems
Verified
Journal Article
Closed access
ANAlexander H. Nelson
Alexander H. Nelson
GSGurashish Singh
Gurashish Singh
RRRyan Robucci
Ryan Robucci
CPChintan Patel
Chintan Patel
NBNilanjan Banerjee
Nilanjan Banerjee
Upper extremity mobility impairment is a common sequel of Spinal Cord Injury (SCI), brain injury, strokes, and degenerative diseases such as Guillain-Barre and ALS. Existing assistive technology solutions that provide access as user input devices are intrusive and expensive, and require physical contact that can have deleterious effects such as skin friction injury for paralyzed users who have reduced skin sensitivity. To address this problem, in this paper, we present the design, implementation, and evaluation of a non-contact proximity gesture recognition system using fabric capacitive sensor arrays. The fabric sensors are lightweight, flexible, and can be easily integrated into items of quotidian use such as clothing, bed sheets, and pillow covers. Our gesture recognition algorithm builds on two known classification techniques, Hidden Markov Model and Dynamic Time Warping to convert raw capacitance values to alphanumeric gestures. Our system is personalized to the user, allowing personalized selection of gesture sets and definition of gesture patterns in accordance with their capabilities. Our system adapts to changes in sensor configuration and orientation with minimal user training and intervention. We have evaluated our system in the context of a gesture-driven home automation system on six subjects that includes an individual who has a C6 Spinal Cord injury. We show that our system can recognize gestures of varying complexity with an average accuracy of 99 percent with minimal training.

17 April 2019 in IEEE Internet of Things Journal
Verified
Journal Article
Closed access
ANAlexander H. Nelson
Alexander H. Nelson
GTGreg Toth
Greg Toth
DLDennis Linders
Dennis Linders
CNCuong Nguyen
Cuong Nguyen
SRSokwoo Rhee
Sokwoo Rhee
Recent efforts to integrate Internet of Things (IoT) technologies into Smart City constructs have proven successful through proof-of-concept deployments in multiple cities. To facilitate additional deployments, standardization, and dissemination of best practices for these technologies is critical. To that effort, this paper seeks to establish a framework in a set of IoT hardware, software, and wireless radio components that are flexible to address several use cases and easily replicable to new deployments. Moreover, the framework aims to be standards-based and to use best practices for cybersecurity to maintain security and privacy. The framework is developed, deployed, and tested through three subprojects in different domains and a live testbed housed by a municipal government, collecting over 160 million distinct sensor records from eight physical locations over a period of eight months. Additionally, the framework is replicated to a separate jurisdiction to prove ease of replicability. This paper identifies common barriers to IoT adoption and replication, both technological and organizational, and enumerates the approaches that were taken to overcome those barriers in these projects.

7 January 2022 in IEEE Sensors Journal
Verified
Journal Article
Closed access
HLHaoyan Liu
Haoyan Liu
APAtiyehsadat Panahi
Atiyehsadat Panahi
DADavid Andrews
David Andrews
ANAlexander H. Nelson
Alexander H. Nelson
Upper-Extremity motor impairment affects millions of Americans due to cerebrovascular incidents, spinal cord injuries, or brain trauma. Current therapy practices used to assist these individuals in regaining motor functionality often require extensive time at rehabilitation facilities with potentially prohibitive travel or financial costs. This work presents a mobile low-cost field programmable gate array (FPGA)-smart rehabilitation system that can be used in home environments. The prototype is a rehabilitation table instrumented with a capacitive sensor array (CSA) to track upper-extremity motions of the user through proximity or touch. In addition, inertial measurement units (IMUs) are placed on the affected upper limb and combined with the CSA data with our sensor fusion signal processing architecture. Motions are classified and evaluated using multi-task convolutional recurrent neural networks with three additional motion quality output classes to personalize recognition based on the particular motor skills of each patient. The prototype achieves above 99% accuracy with 32-bit fixed-point format implementation for recognizing dynamic motions and identifying unnatural characteristics (i.e., tremor or limited flexion and extension) in upper limb motions based on sensor values. The convolutional recurrent neural network (C-RNN) fusion classification network is implemented on a 200 MHz Zynq ZCU104 FPGA using an HLS-based design optimized with pipelining and parallelism techniques and achieves 5.4x speedup compared to ARM® Cortex-A53 implementation running at an operating frequency of 1.3 GHz. The prototype is also demonstrated to perform the machine learning classification in real-time.

16 February 2016 in IEEE Sensors Journal
Verified
Journal Article
Open access
GSGurashish Singh
Gurashish Singh
ANAlexander H. Nelson
Alexander H. Nelson
SLSheung Lu
Sheung Lu
RRRyan Robucci
Ryan Robucci
CPChintan Patel
Chintan Patel
NBNilanjan Banerjee
Nilanjan Banerjee
Individuals with mobility impairments represent a large portion of the population. These impairments are often the sequels of strokes, spinal cord injury, or diseases, such as amyotrophic lateral sclerosis and Guillain-Barré syndrome. Many of these diagnoses manifest in upper-extremity mobility impairment, which may prohibit or make difficult actuating devices within the home. While solutions exist for smart-home automation and control, a few are approachable by those with mobility impairments. To address this issue, we designed Inviz, a touchless low-power assistive gesture recognition system that utilizes an array of textile capacitive sensors to control a smart-home automation system. We present two novel research contributions; the use of flexible textile-based capacitive arrays as differential proximity sensors, and a low-power event-driven hierarchical signal processing algorithm for capacitor sensor arrays. A fully functional prototype system is implemented and evaluated in the context of an end-to-end home automation system using intuitive swipe and hover gestures that are comfortable and approachable for users with mobility impairments. The evaluation demonstrates that the system achieves high accuracy (>90%) while maintaining low latency (<;1 s) and low energy consumption (2475 μW). In addition, we evaluate the system on a subject with upper extremity mobility impairment to verify its usage as an accessibility device.

13 February 2025 in IEEE Access
Verified
Journal Article
Open access
NCNaga Venkata Sai Raviteja Chappa
Naga Venkata Sai Raviteja Chappa
PNPha Nguyen
Pha Nguyen
ANAlexander H. Nelson
Alexander H. Nelson
Han-Seok SeoHan-Seok Seo
Han-Seok Seo
XLXin Li
Xin Li
KLKhoa Luu
Khoa Luu
Social group activity recognition is crucial for various applications including surveillance, human-robot interaction, and behavioral analysis. Current approaches often require extensive manual annotations and rely heavily on pre-trained detectors, limiting their practical applications. Additionally, existing methods struggle to effectively model long-term spatiotemporal relationships in group activities. This paper introduces a novel approach to Social Group Activity Recognition (SoGAR) using Self-supervised Transformers network that can effectively utilize unlabeled video data. To extract spatio-temporal information, we create local and global views with varying frame rates. Our self-supervised objective ensures that features extracted from contrasting views of the same video are consistent across spatio-temporal domains. Our proposed approach efficiently uses transformer-based encoders to alleviate the weakly supervised setting of group activity recognition. By leveraging the benefits of transformer models, our approach can model long-term relationships along spatio-temporal dimensions. Our proposed SoGAR method achieves state-of-the-art results on three group activity recognition benchmarks, namely JRDB-PAR, NBA, and Volleyball datasets, surpassing the current state-of-the-art in terms of F1-score, MCA, and MPCA metrics.