SSD

SSD

Adaptive Researcher

Posts by Collection

portfolio

publications

GENEVAC: Evolving Exits for Evacuation Models via Steady-State Genetic Search

Published in Auburn University CSSE Technical Reports 05-06, 2005

We created the GeneNetwork Question Answer system (GNQA), a generative pre-trained transformer (GPT) knowledge base driven by a performant retrieval augmented generation (RAG) with a focus on aging, dementia, Alzheimer’s, and diabetes.

Recommended citation: Carnahan, B., Muhdi, R., Fu, S.-G., Darnell, S., Davis, J., Dozier, G., Smith, A. (2005). “GENEVAC: Evolving Exits for Evacuation Models Via Steady-State Genetic Search,” Technical Report: Auburn University CSSE05-06, May 18, 2005.
Download Paper | Download Slides | Download Bibtex

Alleviating Escape Panic using Evolutionary Intelligence

Published in Auburn University Electronic Theses and Dissertations, 2006

Reducing damage, danger, and panic by evolving room designs is possible with artificial intelligence. Escape panic, brought on by groups of people being in a life-threatening situation, increases the fatality rate and level of property damage incurred during unfortunate disasters. Currently buildings are designed to a code that tells how many exits a room should possess, but the code does not specify where to place the doors and exactly how many doors there should be in room designs to help alleviate damages to people and property. A genetic algorithm and particle swarm optimizer will be used to find a room design to help alleviate this problem.

Recommended citation: S. S. Darnell, “Alleviating escape panic using evolutionary intelligence,” Auburn University Theses and Dissertations, May 2006. [Online]. Available: http://etd.auburn.edu/handle/10415/495.
Download Paper | Download Slides | Download Bibtex

MyDash: The biometric digital dashboard

Published in International Conference on Intelligent User Interfaces (IUI 2011) 3rd Workshop on Multimodal Interfaces for Automotive Applications (MIAA 2011), 2011

This concept paper introduces a digital dashboard that utilizes biometric identification. The system is designed to have any driver’s personal settings stored using their biometric identity and have the car load personalized settings once it recognizes the authorized biometric signature. The paper evaluates the biometric solutions to be applied in an automotive setup and chooses the well-proven and most ubiquitous, fingerprint. Personalized settings are stored on a secure server over the network and can only be retrieved and unencrypted using the driver’s personal biometric signature.

Recommended citation: Shelby Solomon Darnell, I. Alvarez, J. J. Ekandem, D. L. Woodard, and J. E. Gilbert, “Mydash: The biometric digital dashboard,” in Proceedings of the 3rd Workshop on Multimodal Interfaces for Automotive Applications (MIAA’11) organized at the 2011 International Conference on Intelligent User Interfaces (IUI’11), Palo Alto, CA, Feb. 2011, pp. 53–56.
Download Paper | Download Slides | Download Bibtex

Accessible voting: one machine, one vote for everyone

Published in CHI`11 Extended Abstracts, 2011

The video for Accessible Voting shows a novel technology that allows private and secure voting to people with disabilities who have previously not had the same access to voting equipment as the common voter.

Recommended citation: J. Gilbert, W. Johnson-Stokes, Shelby Solomon Darnell, J. Ekandem, H. Alnizami, and A. M. Martin, “Accessible voting: One machine, one vote for everyone,” in ACM CHI ’11 Video Presentation, Vancouver, Canada, May 2011. [Online]. Available: https://dl.acm.org/citation.cfm?id=1979549.
Download Paper | Download Slides | Download Bibtex

Locus of control in coversational agent design: Effects on older users’ interactivity and social presence

Published in Twelth Interational Conference on Intelligent Virtual Agents (IVA 2012), 2012

This study examines the impact of locus of control in conversational agent design on the interaction experience of 61 older users (65+ years old) through a laboratory experiment. Results reveal that for older users, agent control facilitates greater interactivity and social presence than user control.

Recommended citation: V. Chattaraman, W. Kwon, J. Gilbert, and S. S. Darnell, “Locus of control in conversational agent design: Effects on older users interactivity and social presence,” in Twelth International Conference on Intelligent Virtual Agents (IVA 2012), Santa Cruz, CA, Sep. 2012.
Download Paper | Download Slides | Download Bibtex

Human and machine performance on periocular biometrics under near-infrared light and visible light

Published in IEEE Transactions on Information Forensics and Security, 2012

Periocular biometrics is the recognition of individuals based on the appearance of the region around the eye. Periocular recognition may be useful in applications where it is difficult to obtain a clear picture of an iris for iris biometrics, or a complete picture of a face for face biometrics. Previous periocular research has used either visible-light (VL) or near-infrared (NIR) light images, but no prior research has directly compared the two illuminations using images with similar resolution. We conducted an experiment in which volunteers were asked to compare pairs of periocular images. Some pairs showed images taken in VL, and some showed images taken in NIR light. Participants labeled each pair as belonging to the same person or to different people. Untrained participants with limited viewing times correctly classified VL image pairs with 88% accuracy, and NIR image pairs with 79% accuracy. For comparison, we presented pairs of iris images from the same subjects. In addition, we investigated differences between performance on light and dark eyes and relative helpfulness of various features in the periocular region under different illuminations. We calculated performance of three computer algorithms on the periocular images. Performance for humans and computers was similar.

Recommended citation:
Download Paper | Download Slides | Download Bibtex

Understanding Privacy and Trust Isues in a Classroom Affective Computing System Deployment

Published in Distributed, Ambient and Pervasive Interactions Conference, 2013

Our research group is in the midst of working with teachers to co-design an affective computing system that uses physiological measures, gathered via wrist worn sensors, to understand how students are engaging with classroom instruction. Optimally, our goal is to find new ways of supporting empathetic practices in the classroom by providing teachers real-time (or reflective) feedback on student engagement. In parallel, with our work with teachers, we are working to pinpoint the privacy and trust issues that might be associated with this type of system. The objective of this paper is to present the results of a series of studies conducted to understand the challenges associated with introducing a pervasive affective computing system into classroom environments. While we focus on physiological sensors, the implications apply to other pervasive technologies as well.

Recommended citation: S. Daily, D. Myers, Shelby Solomon Darnell, T. Roy, and M. James “Understanding privacy and trust issues in a classroom affective computing system deployment,” in Distributed, Ambient, and Pervasive Interactions, Springer Berlin Heidelberg, 2013, pp. 419--423.
Download Paper | Download Slides | Download Bibtex

EngageME: A tool to simplify the conveyance of complicated data

Published in Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, 2014

EngageMe is a visualization tool geared towards supporting teachers in understanding how they are connecting with their students and how their pedagogical strategies can be modified to meet the individual needs of a diverse student population. The goal of this application is two-fold: 1) To support instructors in their efforts to create the best possible learning environment for each student, and 2) To provide a tool with which to experiment with different pedagogical strategies. In this abstract, I describe the collaborative and iterative process of developing the graphical interface for this tool with middle school educators before detailing my plans for moving this research forward.

Recommended citation: S. S. Darnell, “Engageme: A tool to simplify the conveyance of complicated data,” in CHI ’14 Extended Abstracts on Human Factors in Computing Systems, Toronto, Canada, 2014. [Online]. Available: https://dl.acm.org/citation.cfm?id=2574765.
Download Paper | Download Slides | Download Bibtex

Human-Computer Interfaces for Speech Applications

Published in CRC Press -> Computing Handbook: computer science and software engineering, 2014

Recommended citation: Shelby Solomon Darnell, N. Mack, H. Alnizami, et al., Computing Handbook, 3rd ed. T. Gonzalez, J. Diaz-Herrera, and A. Tucker, Eds. CRC Press, 2014, ch. Human-Computer Interfaces for Speech Applications, 92:1–92:15. [Online]. Available: https://www.taylorfrancis.com/books/9781439898536/chapters/10.1201%5C%2Fb16812-10
Download Paper | Download Slides | Download Bibtex

EngageMe: Designing a Visualization Tool Utilizing Physiological Feedback to Support Instruction

Published in Technology, Instruction, Cognition and Learning, 2015

Because the conditions under which engagement is elicited may vary, EngageMe, a visualization tool whose purpose is to assist instructors efforts to understand student engagement in the learning process, has been developed.

Recommended citation: M.T. James, S.B. Daily, T. Roy, and S.S. Darnell, “Engageme: Designing a visualization tool utilizing physiological feedback to support instruction,” Technology, Instruction, Cognition and Learning, vol. 10, 2 2015.
Download Paper | Download Slides | Download Bibtex

Engageme: The design and implementation of a reflective tool for evaluating student engagement

Published in Dissertation, Clemson University, 2015

Recently, there has been a growing push to explore the potential of non-cognitive factors in helping students reach their fullest potential. Engagement, one predictor of student achievement, is such a factor. Because the conditions under which engagement is elicited may vary, EngageMe, a visualization tool, has been developed to assist instructors’ efforts to understand student engagement in the learning process. The application attempts to enhance traditional observation methods by utilizing electrodermal activity, a measure of physiological arousal, as a proximal indicator of engagement. An iterative, participatory design process was used to create prototypes of the EngageMe interface. The results of this design process, a study focused on the barriers to adoption of this kind of technology, as well as an exploratory case study are discussed. Finally, implications for future development are presented.

Recommended citation: Shelby Solomon Darnell, “Engageme: The design and implementation of a reflective tool for evaluating student engagement,” Clemson University All Dissertations, vol. 2015, 1790 May 2015.
Download Paper | Download Slides | Download Bibtex

Chapter 9 - Affective Computing: Historical Foundations, Current Applications, and Future Trends

Published in Science Direct, 2017

Affective computing relates to, arises from, or intentionally influences emotion.

Recommended citation: S. B. Daily, M. T. James, D. Cherry, et al., Chapter 9 - Affective Computing: Historical Foundations, Current Applications, and Future Trends, M. Jeon, Ed. San Diego: Academic Press, 2017, pp. 213–231, ISBN: 978-0-12-801851-4. [Online]. Available: https://www.sciencedirect.com/science/article/pii/B9780128018514000094
Download Paper | Download Slides | Download Bibtex

Towards blockchain-enabled school information hub

Published in Proceedings of the Ninth International Conference on Information and Communication Technologies and Development, 2017

Several initiatives have been proposed to collect, report, and analyze data about school systems for supporting decision-making. These initiatives rely mostly on self-reported and summarized data collected irregularly and rarely. They also lack a single independent and systematic process to validate the collected data during its entire lifecycle. Furthermore, schools in developing countries still do not maintain complete and up-to-date school records. Due to these and other factors addressing the education challenges in those countries remains a high priority for local and international governments, donor and non-governmental agencies across the world. In this paper, we discuss our initial design, implementation, and evaluation of a blockchain-enabled School Information Hub (SIH) using Kenyas school system as a case study.

Recommended citation: N. Bore, S. Karumba, J. Mutahi, S. S. Darnell, C. Wayua, and K.Weldemariam, “Towards blockchain-enabled school information hub,” in Proceedings of the Ninth International Conference on Information and Communication Technologies and Development, ser. ICTD ’17, Lahore, Pakistan: ACM, 2017, 19:1–19:4, ISBN: 978-1-4503-5277-2. DOI: 10.1145/3136560.3136584.
Download Paper | Download Slides | Download Bibtex

Secure one-time biometric tokens for non-repudiable multi-party transactions

Published in 2017 IEEE Workshop on Information Forensics and Security (WIFS), 2017

Numerous applications require users to interact with a multitude of entities in order to avail a service. In such applications, the identity of the user is typically verified through physical or digital tokens, which are prone to both identity theft (lost or stolen tokens) and repudiation claims by malicious users. The use of biometrics can provide non-repudiability by ensuring that a purchaser is the bonafide user of the service. However, in applications entailing multiple stakeholders, there may be privacy issues with sharing user’s biometrics data. While this concern can be addressed by storing the user’s biometric data on the token itself, strong mechanisms are required to ensure that token is both secure and tamper-proof. In this paper, we propose a single use biometric token that relies on Shamir’s secret sharing algorithm and blockchain technology to ensure that the encrypted biometric template contained in the token is secure, tamper-proof, and any attempt to use the issued token is irrefutably logged to prove subsequently that the user indeed availed the service. We also analyze issues related to the system security, user privacy, and usability of the proposed solution.

Recommended citation: K. Nandakumar, N. Ratha, S. Pankanti, and S. S. Darnell, “Secure one-time biometric tokens for non-repudiable multi-party transactions,” in 9th International Workshop on Information Forensics and Security, ser. WIFS ’17, Rennes, France: IEEE, 2017. [Online]. Available: https://ieeexplore.ieee.org/abstract/document/8267654/.
Download Paper | Download Slides | Download Bibtex

3 Stages of a Pan-African Identity Framework for Establishing Self-Sovereign Identity with Blockchain

Published in Frontiers in Blockchain, 2021

The three stages are defined by systems that allow for biometric/demographic registration (stage 1), interoperability and security hardening (stage 2), and biometric modality data analysis/organization/association (stage 3).

Recommended citation: Darnell SS and Sevilla J (2021) 3 Stages of a Pan-African Identity Framework for Establishing Self-Sovereign Identity With Blockchain. Front. Blockchain 4:631640. doi: 10.3389/fbloc.2021.631640
Download Paper | Download Slides | Download Bibtex

GlaucoRAG: A Retrieval-Augmented Large Language Model for Expert-Level Glaucoma Assessment

Published in medRxiv/PubMed, 2025

This paper about the creation of a ophthalmologists clinical support tool using LLMs

Recommended citation: Aminan M, Darnell SS, Delsoz M, Nabavi A, Wright C, Jerkins B, Yousefi S. GlaucoRAG: A Retrieval-Augmented Large Language Model for Expert-Level Glaucoma Assessment. [preprint] medRxiv. 2025 Jul 7:2025.07.03.25330805. doi: 10.1101/2025.07.03.25330805. PMID: 40672509; PMCID: PMC12265780.
Download Paper

Creating a biomedical knowledge base by addressing GPT inaccurate responses and benchmarking context

Published in Open Research Africa, 2025

We created the GeneNetwork Question Answer system (GNQA), a generative pre-trained transformer (GPT) knowledge base driven by a performant retrieval augmented generation (RAG) with a focus on aging, dementia, Alzheimer’s, and diabetes.

Recommended citation: Darnell SS, Overall RW, Guarracino A et al. Creating a biomedical knowledge base by addressing GPT inaccurate responses and benchmarking context. Open Res Africa 2025, 8:12 (https://doi.org/10.12688/openresafrica.15854.1)
Download Paper | Download Slides | Download Bibtex

Pangenome-informed Language Models for Synthetic Genome Sequence Generation

Published in 2025 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), 2026

We introduce three innovative pangenome-based tokenization schemes that enhance DNA sequence generation.

Recommended citation: P. Huang, F. Charton, J.M. Schmelzle, S.S. Darnell, P. Prins and E. Garrison, "Pangenome-Informed Language Models for Synthetic Genome Sequence Generation," 2025 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), Wuhan, China, 2025, pp. 5025-5033, doi: 10.1109/BIBM66473.2025.11356310.
Download Paper | Download Slides | Download Bibtex

talks

An Overview of Computer Vision

Published:

As a research scientist with a background in biometrics, pattern recognition and computer vision I gave an invited talk on computer vision at Strathmore University, while fielding questions about pursuing computer science study beyond their undergraduate program.

Metta Blockchain Panel discussion

Published:

During 2016-2018 IBM Reserach made blockchain a technical focal point. As a new hire I picked up blockchain technology and became an expert at its ideation and implementation. At this point I already wrote three patents on blockchain and two research articles.

Blockchain: What, Why & Good to Know

Published:

Upon receiving an invitation to be a guest lecturer for a senior level class on Financial Economics in the Math Department at Strathmore University, I was asked to give a talk on Blockchain and Cryptocurrency. At the time “crypto” was being pursued by many Kenyan youth and few understood its underlying technology or economic principles and the implications.

Blockchain: What, Why & Good to Know

Published:

Moringa School, a cutting edge tech school in Kenya, invited me to give a talk on Blockchain and Cryptocurrency. At the time “crypto” was being pursued by many Kenyan youth and few understood its underlying technology or economic principles and the implications.

Published:

Keynote: Implementing with the Hyperledger Ecosystem

Published:

After demoing a few hyperledger projects, I gave a keynote on implementing with the hyperledger ecosystem. Then I fielded questions about blockchain technology and how the Hyperledger ecosystem can be used to build enterprise solutions.

Professor S on the Role of AI and Blockchain in identities, financial inclusion and governance

Published:

Kenya has been undergoing a digital trust revolution at a break neck speeed. Due to the near ubiquity of mobile devices along with mobile money Kenya’s youth are leading the pushes for financial inclusion and identity management. The current best way to manage ones digital identity is thru a well thought out blockchain solution. The current best way to ensure proper financial inclusion and governance is to broadly use AI in concert with blockchain to support safe digital transactions without trust.

teaching

Fall 2003 Basic Computing Skills

Undergraduate course, Auburn University, Computer Science and Software Engineering, 2003

Taught basic computer usage skills to undergraduates, specifically the Microsoft Office suite of tools.

Teaching experience 1

Undergraduate course, University 1, Department, 2014

This is a description of a teaching experience. You can use markdown like any other post.

Teaching experience 2

Workshop, University 1, Department, 2015

This is a description of a teaching experience. You can use markdown like any other post.