Research Projects

Reimagining the JFK Assassination in the Smartphone Era: AI-Driven Analysis of Historical Footage and Crowd Dynamics

Faculty Member(s) Working on Project: Md Baharul Islam, Ph.D. (PI); Judd Cribbs (Co-PI); Sudip Dhakal, Ph.D. (Co-PI)

Student(s) Working on Project: Sahbah Gonzalez

Description: The assassination of President John F. Kennedy in 1963 remains one of the most visually iconic and publicly scrutinized events of the 20th century. While the Zapruder film provides the only continuous record of the assassination, additional amateur films exist but suffer from distance, grain, shakiness, and obscured sightlines. In the contemporary digital era, the same event would likely be captured by hundreds of smartphones and instantly disseminated across social media platforms. This project proposes to merge computer vision and AI techniques with media studies to reimagine how the Kennedy assassination might have been recorded today. The research will (a) enhance poor-quality archival films, (b) apply AI to detect and count spectators frame by frame, (c) analyze crowd distribution in Dealey Plaza, and (d) model how many smartphone recordings would exist in today’s media environment. The outcomes will provide both technical innovations in archival video restoration and scholarly insights into how technology alters the public witnessing of history.

Funding Source: WCE HeadStart Grant Program

Funding Amount: $5,000

Outside Organizations Collaborating on Project: Department of Communication and Philosophy, FGCU

SignifyAR: An Augmented Reality-Based Platform for Real-Time Multilingual Sign Language Interpretation

Faculty Member(s) Working on Project: Md Baharul Islam, Ph.D.

Student(s) Working on Project: Saif Simanta

Description: The proposed project, SignifyAR, aims to design and prototype an innovative, AI-enabled Augmented Reality (AR) system that facilitates real-time multilingual sign language interpretation. This project addresses a critical gap in inclusive communication technologies by integrating computer vision, deep learning, and AR interfaces to support seamless interaction between signers and non-signers. The system will capture sign language gestures through a camera, process them using deep learning models, and render interpretations using animated avatars or textual overlays via an AR headset or mobile device. This collaborative, faculty-led research initiative will engage undergraduate student researchers in both the software engineering and cognitive interface design components of the system. The interdisciplinary team will include collaborators from Computing and Software Engineering, Psychology, and external partners from the Sally J. Pimentel Deaf and Hard-of-Hearing Center in Fort Myers. Students will gain hands-on experience in dataset annotation, model training, AR integration, and usability testing. The expected outcomes of this 17-month project include (1) development of a high-fidelity prototype of SignifyAR, (2) submission of at least one peer-reviewed conference paper and one journal article, and (3) preparation of a competitive external funding proposal targeting the NSF Smart and Connected Communities or IUSE program. The Holmes Fund award will support student stipends, hardware for AR testing, and travel to disseminate research findings.

Funding Source: Whitaker Institute for STEM Education (Holmes Development Fund)

Funding Amount: $10,000

Currently Hiring

Enhanced Perception to Overcome Visual Field Defects: A Multi-Modal AR/MR-based Visual Aid

Faculty Member(s) Working on Project: Md Baharul Islam, Ph.D.

Student(s) Working on Project: Jack Leser, Ian Wong

Description: The ultimate goal of this project is to break new ground in HMD-based visual aids by developing an XR-based system to maximize the perception of people with VFL (moderate to severe) from their surroundings.

Funding Source: WiSER RA and WCE GA

Funding Amount: Two assistantships

Head-Start: Quantum Reinforcement Learning for Real-Time UAV-Aided Post-Disaster Transportation Infrastructure Resilience

Faculty Member(s) Working on Project: Chengyi Qu, Ph.D. (PI) and Jieyi Bao, Ph.D. (Co-PI)

Student(s) Working on Project: Gabriella Vallar, Ishita Chakkalakkal

Description: This proposal seeks to develop an integrated quantum reinforcement learning (QRL) based framework for real-time UAV-aided roadway and pavement reconstruction after hurricanes. The proposed research builds upon preliminary work in which Generative AI (GenAI) has been employed to synthesize diverse pre-disaster scenarios, enabling models to be pre-trained on a wide range of possible damage conditions. The central challenge addressed here is that post-hurricane roadway assessment and reconstruction planning involve complex decision-making in dynamic, large-scale environments with real-time constraints. Formally, the optimization problem underlying UAV path planning, dynamic data processing, and structural modeling is NP-hard due to its combinatorial nature, the evolving conditions of disaster sites, and the exponential growth in the state-action space when classical reinforcement learning is applied. By introducing QRL techniques, this project aims to reduce the computational complexity from exponential to near-polynomial levels, thereby enabling real-time decision-making for large-scale post-disaster reconstruction efforts. Also, UAV-collected data and QRL-based reconstruction strategies will be integrated into Building Information Modeling/Geographic Information Systems (BIM/GIS) platforms to support intuitive visualization and management, serving as a practical prototype toward future digital twin development.

Funding Source: WCE HeadStart Grant Program

Funding Amount: $5,000

Currently Hiring

Hierarchical AI-Driven and Multimodal Security Framework for Smart Homes in Senior Communities

Faculty Member(s) Working on Project: Chengyi Qu, Ph.D. 

Student(s) Working on Project: Sean Peppers, Oliver Ostlander

Description: Beyond technical advancements, this project will have a transformative impact on digital security for aging communities by establishing a long-term cybersecurity research and education platform in collaboration with the Shady Rest Institute of Positive Aging and other senior-focused organizations. This platform will provide seniors with interactive cybersecurity workshops, incorporating hands-on experiences with smart home security tools, real-time threat awareness demonstrations, and personalized cybersecurity guidance based on multimodal behavioral data. By directly involving seniors in cybersecurity training, the project aims to bridge the gap between advanced security research and real-world adoption. Additionally, this project will integrate multimodal AI-driven cybersecurity concepts into undergraduate and graduate curricula, allowing students to engage in interdisciplinary research that combines IoT security, AI-driven multimodal analysis, and human-centered cybersecurity design. To further extend its outreach, the project will also develop educational modules for K-12 students, introducing fundamental smart home security and AI concepts through interactive learning experiences that emphasize privacy, security, and AI ethics. By combining research, education, and community engagement, this project will not only deliver cutting-edge security solutions for smart homes but also establish a scalable, sustainable model for AI-driven security education. Ultimately, by enhancing cybersecurity awareness and protection in aging populations while cultivating the next generation of security experts, this project will contribute to a safer and more intelligent senior-friendly digital ecosystem.

Funding Source: WiSER Program

Outside Organizations Collaborating on Project: Shady Rest Institute of Positive Aging 

Currently Hiring

Multimodal Object Detection Using Camera, LiDAR, and Radar for Autonomous Driving

Faculty Member(s) Working on Project: Sudip Dhakal, Ph.D. (Co-PI) and Md Baharul Islam (Co-PI)

Student(s) Working on Project: Sebastian Campos, Ehud Orenstain

Description: Autonomous driving systems require highly reliable perception frameworks to detect and
classify objects under diverse environmental conditions. Current multimodal systems
integrating cameras and LiDAR have advanced the field substantially, but they still face
notable limitations. Cameras are affected by poor lighting and adverse weather, while
LiDAR struggles with sparse point clouds at longer ranges.
Radar sensors offer a complementary solution, with strong resilience to adverse weather,
long-range coverage, and inherent velocity estimation capabilities. Although radar has
been extensively studied in defense applications, its integration into civilian autonomous
driving perception frameworks remains underexplored.
This project seeks to build upon prior multimodal detection work by incorporating radar
as an additional modality into a fusion-based object detection framework, thereby
advancing FGCU’s footprint in cutting-edge research on autonomous systems.

Funding Source: WCE HeadStart Grant Program

Funding Amount: $5,000

Investigate AI Application in Electrical Construction

Faculty Member(s) Working on Project: Andy Chau, Ph.D., P.E.; Anna Koufakou, Ph.D.; Long Nguyen, Ph.D., P.E.

Student(s) Working on Project: John Holik

Description: This project aims to explore the integration of AI within the electrical construction industry, addressing the growing need for intelligent tools that can assist in tasks such as interpreting electrical codes, managing submittals, and cross-checking procedures. Electrical contractors face increasing complexity in their operations, and the ability to leverage AI-driven tools can significantly enhance efficiency and reduce errors. The target audience for this research includes electrical contractor executives, project managers, and industry professionals who stand to benefit from the practical applications of AI in their workflows.

AccessiBIMlity

Faculty Member(s) Working on Project: Marina Figueiredo Muller, Ph.D.

Student Worker(s) Working on Project: Leonardo Bacile

Description: This project aims to propose a method to develop tactile maps for people with blindness or visual impairment. Building information models (BIM) can be used to support this process, since the platform allows for extraction and interoperability with 3d printing formats.

Funding Source: WCE HeadStart Grant Program '22

Funding Amount: $5,000

Outside Organizations Collaborating on Project: First Baptist Church of Curitiba

Currently Hiring

REX - From Paper to BIM

Faculty Member(s) Working on Project: Marina Figueiredo Muller, Ph.D. and Raissa Marchiori, Ph.D.

Description: The evolution of project tools, from the pre-paper era to CAD and BIM systems, leads to the question of whether these concepts are fully absorbed by engineering students. A classroom dynamic was proposed to enable students to interact with the process of transitioning between tools. The goal was to develop an activity to teach project tools. This activity was designed based on bibliographic research about project tools and was validated through a survey completed by groups of participants consisting of undergraduate students (with no prior design knowledge) and graduate professionals in engineering. The results demonstrate that the REX dynamic facilitates the consolidation of the learning process.

Outside Organizations Collaborating on Project: University of Louisiana Monroe - Dr. Clint Martin

Currently Hiring

Uncovering Contributing Factors of Electric Vehicle Crashes Through Advanced Data-Driven Analysis

Faculty Member(s) Working on Project: Jieyi Bao, Ph.D.

Description: The rapid growth of Electric Vehicle (EV) adoption has introduced unique safety challenges due to their greater weight, faster torque response, quieter operation, and risk of post-crash battery fires, which may alter crash dynamics and severity. Studies indicate elevated risks in certain scenarios, particularly from mass-related kinetic energy and thermal runaway hazards, yet most crash analyses focus on conventional vehicles, with EV-specific patterns remaining underexplored. Current databases also provide limited distinction between EVs and internal combustion vehicles, constraining insights into roadways, drivers, and vehicle-related factors. This study applies advanced data-driven analytics to address this gap, uncovering hidden relationships in crash data to improve roadway safety and inform infrastructure planning.

Funding Source: WCE HeadStart Grant Program