A5049779043- Context-Aware Activity Recognition Systems
- Neurological disorders and treatments
- Transcranial Magnetic Stimulation Studies
- Obsessive-Compulsive Spectrum Disorders
- Non-Invasive Vital Sign Monitoring
- Tactile and Sensory Interactions
- Robot Manipulation and Learning
- Advanced Sensor and Energy Harvesting Materials
- Hand Gesture Recognition Systems
- IoT-based Smart Home Systems
- Functional Brain Connectivity Studies
- Autonomous Vehicle Technology and Safety
- Balance, Gait, and Falls Prevention
- Social Robot Interaction and HRI
- Soft Robotics and Applications
- Robotic Mechanisms and Dynamics
- Video Surveillance and Tracking Methods
- Anomaly Detection Techniques and Applications
- Retinal and Macular Surgery
- Robotic Path Planning Algorithms
- Robotics and Automated Systems
- Eating Disorders and Behaviors
- Piezoelectric Actuators and Control
- Human Pose and Action Recognition
- Time Series Analysis and Forecasting
University of Pennsylvania
2021-2025
California University of Pennsylvania
2024
Oregon Health & Science University
2024
Stanford University
2024
Veterans Health Administration
2023
Hospital of the University of Pennsylvania
2023
Pennsylvania Hospital
2023
Korea Advanced Institute of Science and Technology
2014-2021
This article proposes an automatic fall detection method for a wearable device that can promptly alert caregivers when is detected, which could reduce the injuries of elder people. To do this, we propose novel generative adversarial network (GAN-) based using heart rate sensor and accelerometer. Acquiring data compared with normal behavioral be arduous process. Instead, introduce compelling GAN-based anomaly partially surrounded User Initial information features (UI-GAN). Although methods...
Treatment-resistant obsessive-compulsive disorder (OCD) occurs in approximately one-third of OCD patients. Obsessions may fluctuate over time but often occur or worsen the presence internal (emotional state and thoughts) external (visual tactile) triggering stimuli. Obsessive thoughts related compulsive urges (are episodic) so respond well to a time-locked brain stimulation strategy sensitive responsive these symptom fluctuations. Early evidence suggests that neural activity can be captured...
This paper proposes an automatic fall detector in a wearable device that can reduce risks by detecting falls and promptly alerting caregivers. For this purpose, we propose cluster-analysis-based user-adaptive detection using fusion of heart rate sensor accelerometer. The objectives the proposed are to have high accuracy with low-complexity model regardless diverse conditions. To meet objectives, best 13-dimensional feature subset selection. In addition, verify performance increment combining...
The Deep Brain Stimulation (DBS) Think Tank XII was held on August 21st to 23rd. This year we showcased groundbreaking advancements in neuromodulation technology, focusing heavily the novel uses of existing technology as well next-generation technology. Our keynote speaker shared vision using neuro artificial intelligence predict depression brain electrophysiology. Innovative applications are currently being explored stroke, disorders consciousness, and sleep, while established treatments...
A self-powered mechanoreceptor array is demonstrated using four cells for recognition of dynamic touch gestures. Each cell consists a triboelectric nanogenerator (TENG) sensing and bi-stable resistor (biristor) spike encoding. It produces informative signals by force an external encoding the into number spikes. An utilized to monitor various gestures it successfully generated corresponding all To validate practicality array, spiking neural network (SNN), highly attractive power consumption...
Fall detection systems have been proposed to prevent additional injuries following fall accidents. This paper introduces an easily learnable system based on the data of individual patient in a hospital room. The improvement low performance using single accelerometer at wrists and inconvenience sensor attached waist conventional approach was concentrated by integrating heart rate signals acceleration changing location from wrists. As for optimal feature selection, we four-feature vector...
Human emotion recognition is an important factor for social robots. In previous research, recognizers with many modalities have been studied, but there are several problems that make rates lower when a recognizer applied to robot. This paper proposes decision level fusion method takes the outputs of each as input and confirms which combination features achieves highest accuracy. We used EdNet, was developed in KAIST based Convolutional Neural Networks (CNNs), facial expression speech...
This paper proposes a monitoring system to prevent falls from bed. The position of patient on the bed is categorized as stable and unstable. has defined unstable condition situation where lying edge was then observed using thermal imagery camera. We extracted x-, y-axis histograms that can be used feature, this SVM (Support Vector Machine) decide an optimal decision boundary achieved accuracy 99.70%.
Interaction between a human and robot can have positive effects on temperament the development of children also be beneficial during psychotherapy. When finishes task, user may want to give reward by means emotional interaction. Here, we assume that mobile is capable interactions. We design system which interacts with humans through use LeapMotion sensor for elaborate hand gesture tracking. The designed combines modeling From this, express emotions intentions using two hands at same time....
Tendon-sheath mechanisms offer a means for flexible surgical robot to be operated efficiently in restricted environments, example, the long and narrow paths inside human organs. However, nonlinear hysteresis interferes with precise motion control of robot. Generally, characteristics can change due changes cable-related effects shapes, which increase difficulty associated achieving robots. Although several methods have been proposed solve this problem, most these limited performance are...
Abstract Background Intraocular surgery and reconstructive are challenging microsurgery procedures that require two types of motion: precise motion larger motion. To effectively perform the requisite using a robot, it is necessary to develop manipulator can adjust scale between less precise, yet Aims In this paper, we propose novel robot dual delta structure (DDS) mechanically seamlessly Materials & Methods The DDS forms lever mechanism enables scaling at end‐effector platforms. Seamless...
Driving assistance systems (DASs) can be useful to inexperienced drivers. Current DASs are composed of front rear monitoring (FRMSs), lane departure warning (LDWSs), side obstacle (SOWSs), etc. Sometimes, provide unnecessary information when using unprocessed low-level data. Therefore, high-level necessary the driver, need improved. In this paper, we present an intelligent driving robotic agent for safe driving. We recognize seven situations, namely, speed bump, corner, crowded area, uphill,...
Fall from height (FFH) is an accident that leads to fatalities in construction workers, and a major cause of FFH due the improper fastening safety hook harness temporary structure. In this study, we propose new approach for recognizing state based on similarity motion between body. We first assume body will be more similar when fastened part than Under assumption, method measures proposing method, motions are represented through acceleration rotations The magnitude as ordinal variable...
Intracranial neurostimulation is a well-established treatment of neurologic conditions such as drug-resistant epilepsy (DRE) and movement disorders, there emerging evidence for using deep brain stimulation to treat obsessive-compulsive disorder (OCD) depression. Nearly all published reports intracranial have focused on implanting single device condition. The purpose this review was educate neurology clinicians the background literature informing dual 2 comorbid neuropsychiatric OCD, discuss...
Abstract / Summary Substance use disorders (SUDs) are a significant public health concern, with over 30% failing available treatment. Severe SUD is characterized by drug-cue reactivity that predicts treatment-failure. We leveraged this pathophysiological feature to personalize deep brain stimulation (DBS) of the nucleus accumbens region (NAc) in an patient. While DBS lead was externalized for clinical purposes, we administered personalized drug cue-reactivity task while recording NAc...
Touch is regarded as an important channel in human-robot interaction. This paper presents a touch gesture recognition system that can be applied to hard-skinned robots. Related studies have been based on traditional machine learning methods with hand-crafted features make it difficult for developers access optimal they cannot imagine. To prevent this, our proposed uses 1D convolutional neural network (1D CNN) learn from data directly. The classifies four patterns: hit, pat, push, and rub....
The awareness for preserving privacy in in-home monitoring robots is increasing. Although several studies have proposed privacy-preserved robot systems adults, only a limited amount of attention has been paid to research on babies. Like previous studies, thermal infrared image-based methods could ensure babies, yet when existing detection were applied images detect babies and we discovered frequent occurrence misdetection due the presence residual heat marks. In this research, propose...
BackgroundLoss of control (LOC) eating, the subjective sense that one cannot what or how much eats, characterizes binge-eating behaviors pervasive in obesity and related eating disorders. Closed-loop deep-brain stimulation (DBS) for binge should predict LOC trigger an appropriately timed intervention.Objective/hypothesisThis study aimed to identify a sensitive specific biomarker detect onset DBS. We hypothesized changes phase-locking value (PLV) LOC-associated cravings distinguish them from...