Join us for presentations, a keynote speaker, a Q&A session, and a chance to network with faculty and industry professionals

Hosted by College of Engineering and Applied Sciences, Long Island Manufacturing Extension Partnership (LIMEP), and Center of Excellence Wireless and Information Technology (CEWIT)

Register here

Location: Stony Brook University, CEWIT Room 152

Abstract:
People shift their visual attention to gather and prioritize information from their surroundings, helping them navigate complex environments. Understanding these attentional shifts involves decoding the features that guide where attention is directed (spatial areas of focus) and when attention shifts (timing). Decoding these processes can aid applications from interface design to medical diagnosis. However, prior models have not fully explored the underlying factors addressing these aspects. In this dissertation, we study the factors that guide visual attention across diverse image types, spanning natural images, graphic design documents, and whole slide images (WSIs) of cancer tissues, while also predicting visual attention based on these factors.
First, we propose a method to quantify object recognition uncertainty as a factor influencing spatio-temporal attention (where and when) in natural images. We found that it plays a larger role than bottom-up saliency in guiding visual attention. Second, we analyze graphic design documents such as webpages, comics, posters, mobile UIs, etc., which differ from natural images in that they are designed to convey specific messages or elicit desired viewer response. We propose a unified and interpretable deep learning model that predicts both static and dynamic visual attention behavior (addressing where and when) by integrating document layout and content saliency as factors, enhancing attention prediction performance. Finally, in the domain of digital pathology, we investigate pathologists' attention during their examination of giga-pixel WSIs of prostate cancer with an objective to aid in the development of computer-assisted pathology training and clinical decision support systems. Using a digital microscope interface, we collected the largest known dataset of pathologist attention, which allows us to study the factors that guide their spatial and temporal attention patterns (where and when) and develop predictive models. Our study explores key factors guiding their attention, including magnification, slide staining, the nature of the diagnostic task, and their expertise. Motivated by this analysis, we propose deep learning models to solve two tasks: 1) predicting pathologist attention via spatial (heatmaps) and spatio-temporal (scanpaths) models, and 2) inferring pathologist expertise level, both essential technical components towards developing an AI-assisted pathology training pipeline.

Speaker:
Souradeep Chakraborty

Location: New Computer Science Bldg., Room 220

Zoom Link: https://stonybrook.zoom.us/j/9755288447?pwd=TW95T2xqOUZjRnlqcnVFcUQvN0JMdz09
Meeting ID: 975 528 8447
Passcode: 338037

The International Conference on Machine Learning (ICML) is the premier gathering of professionals dedicated to the advancement of the branch of artificial intelligence known as machine learning.

ICML is globally renowned for presenting and publishing cutting-edge research on all aspects of machine learning used in closely related areas like artificial intelligence, statistics and data science, as well as important application areas such as machine vision, computational biology, speech recognition, and robotics.

ICML is one of the fastest growing artificial intelligence conferences in the world. Participants at ICML span a wide range of backgrounds, from academic and industrial researchers, to entrepreneurs and engineers, to graduate students and postdocs.


For more information and registration, visit the official website.

As generative AI (GenAI) continues to reshape the educational landscape, educators must critically examine its implications for course design. How can we adapt our courses to ensure meaningful learning in a post-GenAI world? How can we harness its potential while mitigating risks to student learning? This seminar explores the evolving role of GenAI in higher education, emphasizing learner-centered teaching practices--such as backward design, transparency, and active learning--as essential strategies for navigating both the opportunities and challenges posed by GenAI. We will examine how GenAI disrupts traditional models of teaching and assessment, highlighting course design choices that intentionally promote deep learning and critical thinking in this new era.

Speaker Bio: Dr. Lourdes Alemán is an Associate Director at MIT's Teaching and Learning Lab (TLL). She earned her Ph.D. in Biology from MIT, studying RNA interference (RNAi) with Professor Phil Sharp. She later completed a postdoc in curriculum innovation with Professor Graham Walker's HHMI MIT Education Group. As a postdoc and research scientist, she helped develop software tools for teaching experimental design and data analysis, including collaborations with the MIT-Haiti Initiative. Before joining TLL, she worked at MIT's Open Learning, supporting MIT faculty in blended and online education. At TLL, Lourdes trains graduate students and postdocs in college-level teaching, advises faculty on classroom innovation, and previously designed and taught a hands-on biology module on novel antibiotic discovery for first-year students. She has served on university committees focused on mentoring and advising. Drawing from her experiences as a Cuban immigrant student, she developed MIT's first curriculum on growth mindset and co-founded Flipping Failure, a campus-wide initiative for students to share their stories of academic challenges and the strategies they have used to overcome them.

Stony Brook University Northern California Alumni Chapter - Institute for AI-Driven Discovery and Innovation Panel

Join us for a Northern California Alumni and Friends luncheon followed by a panel discussion, celebrating the Institute for AI-Driven Discovery and Innovation, moderated by Fotis Sotiropoulos, Dean, College of Engineering and Applied Sciences.

Panel Discussion with:
Richard Bravman '78, Chief Strategy Officer, Affinity Solutions
Jalal Mahmud, PhD '08, Master Inventor, IBM Watson
Reza Raji '86, CEO, Xenio Systems
Andrew Protter, PhD '83, Chief Scientific Officer, Auansa Inc.

Moderated by:
Fotis Sotiropoulos, Dean, College of Engineering and Applied Sciences

Click here for more information and to register.

Learn how SBU staff are using AI.

Over the past year, we've explored the AI tools and services available at Stony Brook. Now it's time to hear from colleagues who are putting these tools to work in their everyday roles. Join Tamara Gregorian, Herb Schramm, and Allie Seal moderated by David Ecker as they share how they use AI in their position, the tasks it helps them accomplish and the lessons they've learned along the way. This session is an opportunity to see practical examples, ask questions, and gather ideas that you can apply in your own work.

Register here.

Abstract: Much like other AI for Science domains, polymer design poses significant challenges. It requires grounding in empirical data and physical laws, precise handling of domain-specific structured representations, and compositional reasoning over multiple interacting constraints--all while working with limited data.

To address these limitations, we introduce PolyBench, a large-scale benchmark comprising over 125K polymer design and analysis tasks grounded in verified experimental and synthetic data. PolyBench includes tasks created from a wide range of data sources and presents diverse structural, property-driven, and synthesis-oriented reasoning problems. Tasks in PolyBench are organized from simple to complex analytical reasoning problems, enabling generalization tests and includes diagnostic probes to evaluate model capabilities. Additionally, to support effective domain alignment, we propose a knowledge-augmented reasoning distillation framework that enriches the dataset with structured chain-of-thought supervision derived from expert-informed reasoning strategies.

Small language models (7B-14B parameters) trained on PolyBench substantially outperform comparably sized baselines and, in many cases, exceed the performance of larger closed-source frontier models on polymer reasoning tasks, while also demonstrating improved transfer to external polymer benchmarks. Last, we conduct a diagnostic study that reveals a compositionality gap: despite strong performance on decomposed sub-questions, models struggle to integrate multiple interacting constraints and intermediate reasoning steps, highlighting fundamental limitations in current scientific language models.

Speaker: Dikshya Mohanty

Location: NCS 115/Online

Zoom: https://stonybrook.zoom.us/j/94746001760?pwd=BCAd8gu7cXLn3PXM6kkbh11V6r0Mr7.1
Meeting ID: 947 4600 1760 Passcode: 987917


Abstract: Generating high-fidelity EEG data at scale is essential for overcoming dataset scale limitation and satisfying privacy requirements in computational neuroscience. However, the predominant reliance on discrete denoising formulations fails to adequately capture the continuous temporal evolution and frequency-domain characteristics intrinsic to EEG recordings. Consequently, such approaches often compromise long-range temporal coherence and introduce structural discrepancies in both spectral and temporal domains.
In this work, We argue that faithful EEG synthesis demands generative models that directly characterize the continuous dynamics underlying neural signals. To this end, we propose a conditional flow matching framework that treats EEG as raw waveform sequences traversing continuous-time trajectories. Rather than relying on discretized denoising steps or handcrafted signal representations, our method learns a smooth velocity field mapping noise distributions to the target EEG manifold, thereby naturally preserving temporal continuity and transient neural phenomena. To enforce fidelity to fundamental EEG characteristics, we incorporate principled constraint terms that maintain spectral consistency, temporal stationarity, and signal-level statistical properties. Evaluated on large-scale benchmarks, our approach establishes new state-of-the-art results compared to competitive baselines. Comprehensive analyses further confirm that the proposed framework faithfully recovers core structural attributes of neural dynamics, offering a scalable and theoretically grounded solution for high-fidelity EEG generation.

Speaker: Yifan Wang

Location: https://stonybrook.zoom.us/j/95300279172?pwd=maxkSIma0Go2O8Wzre3CLYWHflnUi3.1&jst=2
Meeting ID: 953 0027 9172
Passcode: 418725