Seminar Schedule

Fall 2026

The fall seminar schedule is still being determined.  This page will be updated as information becomes available.

September 3 — Dr. Charlotte kiesel

Designing Audiovisual Pronunciation Feedback for Second Language Learners

As learners acquire a second language, one of the key challenges to developing new phonetic categories (i.e., pronunciation norms) is overcoming the bias created by the learner's first language system. Prior studies show that audiovisual feedback, where learners are provided with a visual representation of their own speech and that of a native speaker, generally outperform traditional audio-only feedback. Though, how learners interpret and integrate feedback remains poorly understood, limiting our ability to design learner-oriented feedback systems.

In this work, we explore and compare learners’ perspectives on these two types of feedback to address this gap. We designed V(is)owel, an audiovisual interactive vowel chart tool, and compared it with audio-only feedback in a within-subjects pronunciation practice study. We observed two key challenges faced by learners: (1) difficulty interpreting feedback, and (2) physically implementing feedback. Our results suggest that visual feedback is effective because it is actionable, provides a goal, and motivates continued practice compared to oral perception alone.

Dr. Charlotte Kiesel is joining Hope College as an Assistant Professor in Computer Science this autumn. She completed her Ph.D. in Computer Science with a focus on Human-Computer Interaction at the University of Illinois Urbana-Champaign in May. In particular, she enjoys researching how students approach foreign language learning and what tools can support them achieve their goals. Beyond academics, she enjoys crafting, playing video games with her husband, and various outdoor activities such as Ultimate Frisbee, beach volleyball, and kayaking.

september 17— claire lundy

Mobile Development: Platforms, Tools, and Real-World Applications

Mobile development combines software engineering with the unique capabilities and constraints of modern mobile devices. This presentation provides a high-level introduction to building mobile applications, covering development tools, iOS and Android platforms, native and cross-platform frameworks, application architecture, navigation, device hardware and sensors, networking, security, and app lifecycle. Through a live demonstration, including an example of a mobile application communicating with a prosthetic device, we’ll explore how mobile software can connect users to real-world hardware and what makes mobile development different from other areas of software development.

Claire Lundy is a software engineer specializing in mobile application development, with experience building applications that connect software with real-world devices and cloud services. She holds a B.S. in Computer Science from Hope College and enjoys exploring the intersection of software, hardware, and accessible technology.

October 1 — redwan sony

Opening the Black Box: Explainability in Biometric Recognition

Biometrics refers to the automated recognition of individuals based on their physical or behavioral characteristics, such as face, fingerprint, iris, voice, and gait. This talk begins with an overview of biometric recognition, including its two principal modes of operation, verification and identification, and the typical pipeline of a biometric system, using face recognition as an example.

Contemporary face recognition systems rely predominantly on domain-
specific recognition models which are essentially deep neural networks
trained on cropped, aligned face images during that achieve remarkable
accuracy but function largely as black boxes: they produce a match score between a pair of face images without any accompanying explanations. In high-stakes applications, the ability to explain a recognition decision is therefore an essential next step. Two broad approaches to produce explanations in face recognition are discussed: saliency-based methods, which identify the facial regions that most influence the match score, and textual explanations, which describe the similarities and differences between a pair of
face images in natural language.

The talk then examines the emerging  use of foundation models, also known as multimodal large language models (MLLMs) trained on large-scale multimodal data, to generate such textual explanations. Although this represents a significant advance toward transparency,  MLLM-generated explanations are prone to hallucination, often referencing facial attributes absent from the images or making claims not grounded in visual evidence.

The talk concludes by presenting ongoing research on evaluating the faithfulness of these explanations at the level of individual claims, with the goal of ensuring that explanations produced by AI systems are not merely plausible, but trustworthy.

Redwan Sony is a fifth-year PhD student in the Department of Computer Science and Engineering at Michigan State University, where he is a member of the iPRoBe Lab advised by Dr. Arun Ross. He holds a bachelor’s degree in computer science and engineering from the Islamic University of Technology (IUT), Dhaka, Bangladesh (2017), and a master’s degree in computer science and engineering from Michigan State University (2023). Before his PhD, he spent three years as a Lecturer in the Department of Computer Science and Engineering at IUT. His research focuses on explainability in face recognition using multimodal large language models, with his dissertation centered on
claim-level faithfulness evaluation of MLLM-generated face recognition
explanations. He has also worked as a Research Scientist Intern at Pindrop Security, focusing on deepfake detection.

October 15 — Alex Atcheson

Generating Access? Investigating the Perspectives of College Students with Disabilities on Generative Artificial Intelligence

In the U.S. alone, it is estimated that nearly one in five college students identifies as having a disability. Yet, many students continue to encounter accessibility barriers that lead to inequities in higher education. Recently, developments in generative artificial intelligence (AI) tools, like ChatGPT, have made possible new forms of interaction and personalization that have applications to accessibility. For example, disabled communities have used generative AI to make creative work more accessible, bridge communication gaps, and help users navigate physical spaces. At the same time, generative AI poses a variety of risks, such as hallucinating information and perpetuating harmful stereotypes. With more students adopting generative AI, our understanding of how this technology may improve (or reduce) the inclusiveness of higher education for students with disabilities is still developing. To what extent are generative AI tools making higher education more accessible to students with disabilities, and what are the potential harms of relying on AI as a solution for addressing disability-related inequities? In this talk, I will share my research on this topic and discuss opportunities for technology’s role in the design of more inclusive learning environments.

Alex Atcheson (https://saatcheson.github.io/) is a sixth-year PhD candidate in Computer Science in the Siebel School of Computing and Data Science at the University of Illinois Urbana-Champaign (UIUC). He received his Bachelor’s degree in Computer Engineering and Computer Science from the University of Southern California in 2021. At UIUC, Alex is advised by Dr. Karrie Karahalios as part of the Social Spaces research group, where he currently studies how AI technologies impact the accessibility of higher education to students with disabilities, collaborating closely with experts in Special Education and Disability Services. In addition to accessibility, education, and AI, Alex's research interests also include governance in online communities, co-design and participatory design methods, and the design of technology to support civic participation.

November 5 -- to be determined

The speaker for this session has not yet been determined

November 19 — Dr. Derek Schuurman

Dr. Schuurman will speak on the topic of Faith and AI.  More details to come.

Dr. Schuurman's personal story includes working in industry for several years before returning to academia. His interest in technology began in his early teens playing with ham radio and his first computer. He is also interested in the bigger picture of ethics, perspectives, and issues related to technology using insights from the philosophy of technology in conversation with the Reformed tradition.

He is author of the book Shaping a Digital World and co-author of A Christian Field Guide to Technology for Engineers and Designers, both published by IVP Academic. He is also a regular contributor to the Christian Scholars Review blog, and a columnist for Christian Courier.

Dr. Schuurman is a fellow of the American Scientific Affiliation, a fellow of the International Society for Science and Religion (FISSR), an associate fellow of the Kirby Liang Centre, and was made a senior member of the IEEE. He is a member of the ACM, CES, ACMS, and serves as a book review editor for Perspectives on Science and Christian Faith.

december 3 - TBD

The speaker for this seminar has not yet been determined.