A5101726686- Combustion and flame dynamics
- T-cell and B-cell Immunology
- Single-cell and spatial transcriptomics
- Video Surveillance and Tracking Methods
- Fluid Dynamics and Turbulent Flows
- Automated Road and Building Extraction
- Mathematical Biology Tumor Growth
- Nuclear and radioactivity studies
- User Authentication and Security Systems
- Aerodynamics and Acoustics in Jet Flows
- Hand Gesture Recognition Systems
- Gait Recognition and Analysis
- Nuclear reactor physics and engineering
- Privacy, Security, and Data Protection
- Fluid Dynamics and Heat Transfer
- Autonomous Vehicle Technology and Safety
- Structural Analysis and Optimization
- Peptidase Inhibition and Analysis
- Robot Manipulation and Learning
- Data Quality and Management
- Offshore Engineering and Technologies
- Robotics and Automated Systems
- Infrared Thermography in Medicine
- Evolutionary Algorithms and Applications
- Melanoma and MAPK Pathways
Oak Ridge National Laboratory
2020-2024
Carnegie Mellon University
2022
City of Hope
2019
Beckman Research Institute
2019
Purdue University West Lafayette
2016-2019
Cloud Computing Center
2018
Applied Materials (United States)
2009
TechnipFMC (Germany)
2004
Face recognition technology has advanced significantly in recent years due largely to the availability of large and increasingly complex training datasets for use deep learning models. These datasets, however, typically comprise images scraped from news sites or social media plat-forms and, therefore, have limited utility more security, forensics, military applications. applications require lower resolution, longer ranges, ele-vated viewpoints. To meet these critical needs, we collected...
We conducted an online survey and remote usability study to explore user needs related advertising controls on Facebook determine how well existing align with these needs. Our results highlight a range of objectives controlling ads, including being able select what ad topics are shown or personal information is used in targeting. Some achievable Facebook's controls, but participants seemed be unaware them, suggesting issues discoverability. In our study, noted areas which the could improved,...
The state-of-the-art in biometric recognition algorithms and operational systems has advanced quickly recent years providing high accuracy robustness more challenging collection environments consumer applications. However, the technology still suffers greatly when applied to non-conventional settings such as those seen performing identification at extreme distances or from elevated cameras on buildings mounted UAVs. This paper summarizes an extension largest dataset currently focused...
I njector spray characteristics have a significant influence on the combustion performance in gas turbine engine, including an impact dynamics, emissions and component life. Furthermore, commercial aviation faces fuel cost, environmental, energy security challenges that arise from use of petroleum based jet fuels. Sustainable alternative fuels can help address these need to be characterized their performance. The present work describes detailed characterization several using hybrid airblast...
We present a cloud-based human-robot interaction system that automatically controls humanoid robot to mirror human demonstrator performing flag semaphores. use framework called Human Augmented Robotic Intelligence (HARI) perform gesture recognition of the and control local robot, named Pepper. To ensure is real-time, we design maximize cloud computation contribution deep-neural-network-based system, OpenPose, minimize communication costs between robot. A hybrid used hide latency caused by...
Evaluating efficiency and security of Connected Autonomous Vehicles (CAVs) requires an environment that can support applications measurements under real-world conditions. This work introduces our implementation evaluation a Vehicle Research Environment (CAVRE). We implement evaluate existing CAV application called Cooperative Adaptive Cruise Control (CACC) using physical Vehicle-to-Vehicle (V2V) communications between virtual agent real autonomous vehicle operating on steerable dynamometer....
This paper examines covariate effects on fused whole body biometrics performance in the IARPA BRIAR dataset, specifically focusing UAV platforms, elevated positions, and distances up to 1000 meters. The dataset includes outdoor videos compared with indoor images controlled gait recordings. Normalized raw fusion scores relate directly predicted false accept rates (FAR), offering an intuitive means for interpreting model results. A linear is developed predict biometric algorithm scores,...
The considerable body of data available for evaluating biometric recognition systems in Research and Development (R\&D) environments has contributed to the increasingly common problem target performance mismatch. Biometric algorithms are frequently tested against that may not reflect real world applications they target. From a Testing Evaluation (T\&E) standpoint, this domain mismatch causes difficulty assessing when improvements State-of-the-Art (SOTA) research actually translate improved...
As the semiconductor industry continues to scale feature sizes, higher deposition/etch rate and better gap-fill abilities are continuously required for chemical vapor deposition (CVD) other processes. To achieve rates, plasma power density required, which can lead arcing. Arcing is a severe issue common CVD processing. Minor arcing generate particles/defects. Severe completely destroy wafers reactor components. In this paper, three typical cases examined using 300mm system. These include...
Abstract While the density of tumor-infiltrating lymphocytes (TILs) is now well known to correlate with clinical outcome, significance spatial distribution TILs not characterized. We have developed novel statistical techniques (including fractal dimension differences, a hotspot analysis, box counting method that we call ‘occupancy’ and way normalize cell ‘thinning’) analyze (at different length scales) various types in triple negative breast tumors. Consistent prior reports, CD20 + B cells...
Machine learning (ML) offers powerful methods for detecting and modeling associations often in data with large feature spaces complex associations. Many useful tools/packages (e.g. scikit-learn) have been developed to make the various elements of handling, processing, modeling, interpretation accessible. However, it is not trivial most investigators assemble these into a rigorous, replicatable, unbiased, effective analysis pipeline. Automated machine (AutoML) seeks address issues by...
We present the design and rationale for user interfaces Privacy Enhancements Android (PE Android). These UIs are built around two core ideas, namely that developers should explicitly declare purpose of why sensitive data is being used, these permission-purpose pairs be split by first party third uses. also a taxonomy purposes ways how ideas can deployed in existing ecosystem.
The goal of immunotherapy is to enhance the ability immune system kill cancer cells. Immunotherapy more effective and, in general, prognosis better, when cells infiltrate tumor. We explore question whether spatial distribution rather than just density tumor important forecasting recurs. After reviewing previous work on this issue, we introduce a novel application maximum entropy quantify discrete point-like objects. apply our approach B and T images tissue taken from triple negative breast...