Brendan Driscoll

North Carolina State University

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Trends

Author Order

Document Type

Co-Authors

Garima Pruthi

Google (United States)

Zhiwei Jia

Taiyuan University of Technology

Kunal Mankodiya

University of Rhode Island

Varun Jampani

Google (United States)

Soravit Changpinyo

Google (United States)

He Huang

University of North Carolina at Chapel Hill

Ming Liu

Huanghuai University

Arjun Akula

IT University of Copenhagen

Suyash Damle

Google (United States)

Pradyumna Narayana

Google (United States)

Cooperative Institutions

University of Rhode Island

Google (United States)

University Health Care System

North Carolina State University

North Central State College

Massachusetts Institute of Technology

Carnegie Mellon University

National Training Service

Indian Institute of Science Bangalore

Georgia Institute of Technology

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Field

MetaCLUE: Towards Comprehensive Visual Metaphors Research

2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (2023)

Arjun Akula Brendan Driscoll Pradyumna Narayana Soravit Changpinyo Zhiwei Jia

Creativity is an indispensable part of human cognition and also an inherent part of how we make sense of the world. Metaphorical abstraction is fundamental in communicating creative ideas through nuanced relationships between abstract concepts such as feelings. While computer vision benchmarks and approaches predominantly focus on understanding and generating literal interpretations of images, metaphorical comprehension of images remains relatively unexplored. Towards this goal, we introduce Meta-CLUE, a set of vision tasks on visual metaphor. We also collect high-quality and rich metaphor annotations (abstract objects, concepts, relationships along with their corresponding object boxes) as there do not exist any datasets that facilitate the evaluation of these tasks. We perform a comprehensive analysis of state-of-the-art models in vision and language based on our annotations, highlighting strengths and weaknesses of current approaches in visual metaphor classification, localization, understanding (retrieval, question answering, captioning) and generation (text-to-image synthesis) tasks. We hope this work provides a concrete step towards developing AI systems with human-like creative capabilities. Project page: https://metaclue.github.io

Closed captioning

Abstraction

Literal (mathematical logic)

Strengths and weaknesses

10.1109/cvpr52729.2023.02222

Cite

Citations (13)

MetaCLUE: Towards Comprehensive Visual Metaphors Research

arXiv (Cornell University) (2022)

Arjun Akula Brendan Driscoll Pradyumna Narayana Soravit Changpinyo Zhiwei Jia

Creativity is an indispensable part of human cognition and also an inherent part of how we make sense of the world. Metaphorical abstraction is fundamental in communicating creative ideas through nuanced relationships between abstract concepts such as feelings. While computer vision benchmarks and approaches predominantly focus on understanding and generating literal interpretations of images, metaphorical comprehension of images remains relatively unexplored. Towards this goal, we introduce MetaCLUE, a set of vision tasks on visual metaphor. We also collect high-quality and rich metaphor annotations (abstract objects, concepts, relationships along with their corresponding object boxes) as there do not exist any datasets that facilitate the evaluation of these tasks. We perform a comprehensive analysis of state-of-the-art models in vision and language based on our annotations, highlighting strengths and weaknesses of current approaches in visual metaphor Classification, Localization, Understanding (retrieval, question answering, captioning) and gEneration (text-to-image synthesis) tasks. We hope this work provides a concrete step towards developing AI systems with human-like creative capabilities.

Closed captioning

Abstraction

Literal (mathematical logic)

Strengths and weaknesses

10.48550/arxiv.2212.09898

Cite

Citations (1)

Design and Characterization of a Helmet-Based Smart Textile Pressure Sensor for Concussion

Arar S. Alkhader Manob Jyoti Saikia Brendan Driscoll Kunal Mankodiya

A Mild traumatic brain injury (mTBI) or concussion has become a public health problem in the United State. Sports and recreational activities are major sources of concussions; with the most incidents connected to American football. Recently, many companies and research institutions have started studying concussions and introduced some means of protection and some alarming systems of strong jolts. The major detection and protection system currently available on the market is the electronic helmet (e-helmet) composed of measurement devices to record head impact acceleration. The most commonly used devices in e-helmets are accelerometers to measure linear acceleration and gyroscopes for rotational/angular acceleration. Using smart textiles for concussion detection is currently uncommon and limited due to the lack of literature studying their voltage related errors. Actually, there are few works that characterize some voltage-force related errors for such type of sensors but for small impact forces and under bench testing while the behavior of those sensors was not described for higher ranges of applied forces and in field situations. This paper previews some common techniques used in e-helmets for concussion detection and highlights electronic textiles and smart fabric sensors that could be very useful for these applications. It discusses and validates the general behavior of such type of sensors under high impact forces and on field testing instead of bench testing, and also it characterizes the effect of increasing the thickness of the sensing element layer on the sensor. A custom-made pressure sensor was created of some available fabrics to be embedded within the padding of a football helmet to quantify the impacting force to the head. The sensor is mainly composed a Semi Conductive Polymer Composite SCPC layer with modifiable thickness that was modified three times with 0.2, 0.4, and 1.6mm to characterize the general behavior of the sensor due to a high amount of impacts and correlated with the thickness. A pendulum system was built to test the pressure sensors, while a special camera and an open-source video analysis software, Tracker was used to track the pendulum bob. The speed and the acceleration of the pendulum bob were measured, then the impact force was calculated and a voltage-force response was obtained. The results showed that no meaningful improvement occurs by small increase in the thickness but better sensor behavior could be obtained by significant increment to observe any difference. Despite that at a very high impacts, the suggested sensor with Velostat layers is not giving the real voltage readings that reflect the actual applied forces but it gives a helpful information that illustrate the distribution of the force through identification the place of the highest and lowest voltage readings regardless of the exact values of those readings. However, the proposed smart textile pressure sensor could be applicable in future e-helmet designs with additional research-based improvements especially on the structure of the sensing element layer to be able to withstand such high impacts which in turns improves the overall sensor performance and accurately measures pressure in concussion-inducing ranges.

10.20944/preprints202007.0629.v1

Cite

Citations (3)

Strongyloidiasis Hyperinfection Syndrome in COVID-19 Positive Migrants Treated with Corticosteroids

Journal of Immigrant and Minority Health (2022)

Nancy Piper Jenks Brendan Driscoll Tiffany Locke

Strongyloidiasis

Strongyloides stercoralis

Strongyloides

Medical record

10.1007/s10903-022-01386-w

Cite

Citations (11)

1-D Manual Tracing Based on a High Density Haptic Stimulation Grid - a Pilot Effort

Brendan Driscoll Ming Liu He Huang

Lower limb amputees lack the neurological path-ways needed for perception of how their prosthetic limbs are interacting with the environment, leading to a lack of confidence in their devices and reduced balancing capabilities. Sensory substitution methods, such as vibrotactile and electrotactile feedback applied to unaffected body segments offer a potential way to restore some of the lost information pathways. While high resolution haptic stimulation grids have become commercially available, few studies have tried to make use of these devices to provide more intuitive sensory substitution methods. This study developed an encoding approach, which is based on the illusory "phantom actuator" phenomenon, to convert 1-D position information to a wearer through a bHaptics Tactsuit. By evaluating performance of 1-D manual tracking task among 14 participants under the proposed approach and a traditional amplitude modulation approach, we demonstrated an improvement of velocity tracing accuracy (p=0.0375) with the proposed approach, although the proposed approach did not lead to significant improvement in the position tracing accuracy.

Tracing

10.1109/whc56415.2023.10224505

Cite

Citations (0)

Nursing Infection Control Practice Adherence, Related Barriers, and Methods of Intervention

JONA The Journal of Nursing Administration (2022)

Brendan Driscoll Darlene Evans

OBJECTIVE This study aims to evaluate the level of infection control (IC) practice adherence and related barriers among nursing staff at a US academic medical center. Furthermore, it surveys staff to identify interventions to improve adherence with IC practices. BACKGROUND Infectious disease emergencies have exposed healthcare workers to increased infection risk. This study aims to explore IC practice adherence in the tertiary hospital setting and further investigate the personal protective equipment (PPE) champion role as a method to boost adherence. METHODS This descriptive quality improvement study includes 2 parts: 1) A PPE champion role was piloted to support healthcare worker adherence to IC practices; and 2) a survey was disseminated to nursing staff to identify the perceptions of their own and coworkers' adherence to IC practices as well as suggestions for improvement. RESULTS Twenty-six percent of PPE champion observations found nursing staff nonadherent with IC practices and policies. Barriers included knowledge deficit (38%), lack of motivation (38%), time constraints (21%), and lack of supplies (3%). Two hundred sixty-four RNs and nurse technicians responded to the IC survey. Nursing staff rated their IC adherence higher than observed by their peers. Perceived barriers to adherence included the lack of supplies (53%), knowledge of proper technique (41%), time constraints (35%), and motivation (5%). Staff recommended using PPE champions (52%) and mandating supervised practice of essential skills (48%) to increase adherence. CONCLUSIONS Nursing staff reported barriers related to their knowledge of proper technique and suggested interventions including supervised practice of critical skills. It seems that the PPE champion role is a popular method to boost IC practice adherence that can be implemented during pandemic and nonpandemic environments at minimal cost.

Champion

Personal Protective Equipment

10.1097/nna.0000000000001120

Cite

Citations (4)

Patient Self-Assessment of 3D Printed Upper-Extremity Prosthetics

Joshua V. Gyllinsky James Gannon Corvah Akoiwala Cristian R. Witcher Laura Camila Hernández Parra

With 40 million amputees living in poor countries with limited access to health care, there is a significant need for novel prosthetic solutions. This paper presents our preliminary results to develop sustainable prostheses for five upper-extremity amputee patients in Colombia, South America. We utilized a patient-centric methodology, engaging patients at every stage in an effort to increase their satisfaction with their prosthetic and to assess clinical usefulness. We identified patient-specific criteria and compared existing open source hardware systems using the Design Thinking methodology. Finally, we present these data to further the development of myoelectric prosthetics as part of cyber-physical systems.

3d printed

10.1109/smartcomp.2019.00054

Cite

Citations (1)

Standing and Walking Attention Visual Field (SWAVF) task: A new method to assess visuospatial attention during walking

Applied Ergonomics (2022)

Jing Yuan Xiaolu Bai Brendan Driscoll Ming Liu He Huang

Sitting

Convergent validity

10.1016/j.apergo.2022.103804

Cite

Citations (2)