HAAG for BioTech

The biotech projects of the Human-Augmented Analytics Group (HAAG) focus on applied data science research tailored specifically for ecological and biological sciences. HAAG employs advanced computational techniques—including machine learning, computer vision, audio analysis, and statistical modeling—to address critical real-world challenges such as wildlife behavior monitoring, species identification, and biodiversity assessments.

HAAG uses cutting-edge data collection technologies like camera traps, drones, acoustic sensors, and LiDAR scans to capture comprehensive ecological data. These tools enable detailed studies of diverse wildlife, including mammals such as deer, birds like leaf warblers, reptiles like anoles and lizards, and various insect species. Projects often involve sophisticated analyses such as tracking animal behaviors, identifying vocalization patterns, and mapping species distribution.

By integrating state-of-the-art vision-language models and large language models (LLMs), HAAG provides robust tools for efficiently processing and interpreting complex datasets. This technological framework supports detailed ecological modeling, behavioral classification, and advanced biodiversity monitoring, directly contributing to innovative ecological research and conservation strategies.

Seminars

HAAG hosts an interdisciplinary seminar series that brings together research, technology, and practical applications in data science and AI. Each semester, we feature industry professionals, academic researchers, and HAAG members to explore topics ranging from AI applications in law and healthcare to federated learning and research methodologies. Our speakers have included engineers from Lockheed Martin presenting bias detection research, law professors examining AI’s role in the U.S. justice system, and academics offering guidance for PhD applicants. These seminars help to foster a learning community where participants can learn about new technologies, engage with the ethical questions surrounding AI, and showcase their own research projects. The series aims to create connections across research groups while also exploring the practical applications of theoretical research. Find out more about past and upcoming seminars by clicking the Seminars button below.

Research Units

Research Units are thematic groups, each encompassing several projects that share similar research goals, tools, or domains. Their primary purpose is to foster collaboration and preserve resources across the research community. Each unit serves as a hub where student researchers, faculty, and experts can exchange knowledge, address technical and strategic challenges, document solutions, and promote best practices.

At HAAG, Research Units hold bi-weekly meetings to encourage the exchange of useful information among researchers with aligned objectives. Each unit is coordinated by a Unit Manager, who is responsible for organizing meetings, maintaining documentation, managing shared resources, and ensuring that contributions are tracked and credited.

Unit meetings help advance research goals through the following processes:

• Presentations, discussions, and code demos led by unit leaders (faculty or computational advisors), invited experts, or student researchers.
• Requests for contributions to the unit’s Zotero literature collection and tools/resources page
• Participation tracking, where students earn credit for literature contributions, problem-solving, demos, and discussions.

To support collaboration and resource sharing, Research Units use the following tools:

• Unit Website: Maintained by the unit and linked to the main HAAG site, featuring project summaries, seminar recordings, and resource links
• Public Zotero Collection: A shared library of literature relevant to the unit’s theme, with contributions from both advisors and researchers
• Tools & Tutorials Page: A collection of non-literature resources (e.g., software tools, walkthroughs), open to contributions from students and advisors
• FAQ Page: A growing archive of technical problems and solutions, sourced from community forums (e.g., Wildlabs, Stack Overflow, GitHub) and unit discussions

This structured approach ensures that knowledge and resources are continuously developed, shared, and applied across related projects.

Faculty Advisors

Dr. Arthur Porto

Faculty Advisor

Dr. Arthur Porto is the Assistant Curator of Artificial Intelligence at the Florida Museum of Natural History. His research integrates artificial intelligence and evolutionary biology, focusing on the development of AI-assisted tools for processing biodiversity data. Dr. Porto has an extensive publication record and has been recognized with numerous grants and awards for his contributions to the fields of AI and evolutionary biology.

Dr. Porto’s work with HAAG leverages a suite of modern AI technologies: from vector‐search pipelines to large language model (LLM) dispatchers that facilitate intuitive, conversational querying of biodiversity databases. On the 3D side, his projects employ photogrammetric reconstruction methods to build detailed models from multi‐angle photos, automated segmentation algorithms to isolate specimen details, and statistical shape models to predict and fit high‐fidelity 3D geometries from partial scans. These technologies enable scalable, high‐throughput analysis of natural history collections and support advanced ecological and evolutionary research.

Dr. Jeffery Cannon

Faculty Advisor

Dr. Jeffery Cannon is a Landscape Ecologist at the Jones Center at Ichauway, where his research program focuses on the causes and consequences of spatial heterogeneity in forested systems across the eastern and western United States. His work examines how natural disturbances such as tornadoes and hurricanes, as well as management actions like prescribed fire, alter the spatial structure of forests and influence fuels, fire behavior, abiotic conditions, and regeneration responses. Dr. Cannon’s current research includes quantifying hurricane risk to forest resources through multi-scale wind risk studies, understanding overstory drivers of forest dynamics in longleaf and related pine systems, and improving restoration outcomes in ponderosa and longleaf pine forests.

Dr. Cannon uses landscape-scale ecological modeling and technology to address real-world problems, and his research has shaped policy decisions, restoration projects, and conservation strategies across the country. Dr. Cannon also engages leaders through refining environmental questions and designing analyses that yield actionable directions for environmental decision-making. Dr. Cannon earned his Ph.D. in Plant Biology from the University of Georgia (2015), following his M.S. in Biology from the University of Mississippi (2011) and B.S. in Biological Sciences from Mississippi State University (2009).

Dr. James Stroud

Faculty Advisor

Dr. James Stroud is an Assistant Professor and an evolutionary ecologist whose research explores the ecological and evolutionary processes that shape biodiversity across space and time. He earned his Ph.D. in Ecology & Evolution from Florida International University and holds a B.Sc. (Hons) in Zoology with Conservation from the University of Wales.

His work focuses on the evolutionary ecology of lizards, using this diverse and dynamic group as a model system to investigate species interactions, community assembly, functional morphology, and the impacts of global change, including climate shifts and biological invasions. His interests span both micro- and macroevolutionary scales, blending natural history with cutting-edge ecological and evolutionary theory.

The Stroud Lab takes a multidisciplinary approach, integrating fieldwork with macroecological and evolutionary analyses, and drawing from fields such as behavior, physiology, biomechanics, and conservation biology. Through this integrative lens, Dr. Stroud aims to uncover the mechanisms that drive patterns of biodiversity and how these patterns respond to environmental change.

He leads a team of researchers committed to advancing our understanding of biodiversity and the complex systems that sustain it.

Dr. Benjamin Freeman

Faculty Advisor

Dr. Freeman is an Assistant Professor and Elizabeth Smithgall-Watts Endowed Faculty in the School of Biological Sciences at Georgia Tech. His Mountain Bird Lab investigates the origin and loss of biodiversity—specifically how climate change drives species loss—by integrating field observations, controlled experiments, and large-scale datasets to map and predict how species’ ranges shift in a warming world.

He earned his B.A. from Macalester College (mentored by Dr. Mark Davis) in 2006, completed his Ph.D. in Ecology and Evolutionary Biology at Cornell University under Dr. John Fitzpatrick in 2016, and then served as a Banting/NSF Postdoctoral Fellow at UBC’s Biodiversity Research Centre.

His Leaf Warbler studies harness cutting-edge computer vision and deep learning: in the Behavior Analysis project he applies motion detection, object tracking, and re-identification models to Himalayan camera-trap footage, then uses supervised classifiers, pose-estimation, and trajectory-analysis algorithms to flag and quantify feeding, predator–prey, and social behaviors against microhabitat variables; concurrently, his Audio Analysis efforts build on the open-source BirdNET CNN framework—enhanced with transfer learning, attention-based architectures, and advanced spectrogram processing—to perform high-accuracy species identification, while LSTM/GRU sequence models and unsupervised clustering map call distributions and test hypotheses about the role of “buzz” vocalizations in mate attraction.

Dr. Jenny L. McGuire

Faculty Advisor

Dr. McGuire’s research explores how terrestrial ecosystems shift in response to changing climates and anthropogenic pressures. She is especially interested in distribution dynamics and uses historical and paleontological data to understand responses to global change.

Dr. McGuire’s work with HAAG integrates advanced computer vision, machine learning, and statistical modeling to address key challenges in ecology and paleontology. She develops algorithms for automated wildlife detection and behavioral analysis from camera trap footage, including high-accuracy identification of species such as moths, deer, and birds, with advanced pose estimation and trajectory analysis. She also leads the development of an R package for analyzing stable isotope data from dental tissues, aimed at improving ecological modeling across temporal gradients through robust statistical methods and uncertainty quantification. Additionally, she directs a project on limestone-fossil differentiation using deep learning to enhance fossil identification accuracy from photographs, contributing tools for both researchers and citizen scientists via accessible platforms.