AI Innovation Institute

Abstract: Visual generation is a fundamental problem in computer vision and graphics, with applications ranging from 3D capture to content creation and image/video synthesis. Despite rapid progress in neural rendering and generative models, efficiency remains a key obstacle in practice: high-quality 3D reconstruction often depends on dense multi-view supervision; scalable 3D synthesis faces heavy optimization, training, and rendering costs; and modern image/video generators incur substantial computation as token grids grow with spatial resolution and temporal length.
This thesis targets efficient visual world modeling by improving sample efficiency in 3D reconstruction, representation efficiency in 3D generation, and computational efficiency in image/video synthesis. First, we improve sample efficiency for neural implicit surface reconstruction under sparse views by integrating multi-view stereo probability volumes as a geometric regularizer, enabling high-quality reconstruction from as few as three input images. Next, we introduce an explicit 3D representation for 3D generation, built from multi-view depth and RGB predictions with 3D Gaussian features, which enables the use of 2D generative priors while enforcing multi-view consistency via epipolar attention. We then address the computational bottleneck of image and video synthesis with importance-based token merging, using importance signals available during generation to preserve critical information while merging redundant tokens. Finally, we propose efficient mixed-resolution diffusion transformers via cross-resolution phase-aligned attention, aiming to improve attention stability under mixed token grids and support high-fidelity mixed-resolution generation.

Speaker: Haoyu Wu

Location: NCS120

Climate Uncertainty, Decision Making, and AI for Earth System Predictability Dr. Nathan Urban, Brookhaven National Laboratory

Bio: Nathan Urban is the group leader of the Optimal Experimental Design & Uncertainty Quantification group in the Applied Mathematics Department at Brookhaven National Laboratory's Computing & Data Sciences directorate (CDS). He holds a Ph.D. in theoretical condensed matter physics from Penn State, and has previously held research positions at Los Alamos National Laboratory, Princeton, and Penn State. His research interests include Bayesian inference and spatiotemporal statistics, probabilistic prediction and forecasting, multi-model / model-form / model structural uncertainty quantification, reduced order modeling, scientific machine learning and hybrid physical-data driven modeling, in-situ/streaming data analysis at scale, information fusion, decision making under uncertainty and optimal experimental design, and integrated multiscale computational frameworks for decision support.

Location: IACS Seminar Room

Lunch will be provided

CG Group member (and SBU faculty) Chao Chen will speak on Fri, March 12, about the use of topological data analysis in machine learning for image analysis.
Chao has shared some of his research with the CG Group previously, and this will be a great opportunity to learn more about this exciting research area related to computational geometry/topology!

Time: Friday, March 12, 2pm-3pm
Place: Zoom
https://stonybrook.zoom.us/my/profweizhu?pwd=RjVIVXg3YUhudzZZQ3pheHUydTJBUT09



Title: Learning with Topological Information - Image Analysis and Label Noise
Speaker: Prof. Chao Chen (SBU)

Abstract: Modern machine learning faces new challenges. We are
analyzing highly complex data with unknown noise. Topology provides
novel structural information to model such data and noise. In this
talk, we discuss two directions in which we are using topological
information in the learning context. In image analysis, we propose a
topological loss to segment and to generate images with not only
per-pixel accuracy, but also topological accuracy. This is necessary
in analysis of images of fine-scale biomedical structures such as
neurons, vessels, etc.  Extracting these structures with correct
topology is essential for the success of downstream
analysis. Meanwhile, we discuss how to use topological information to
train classifiers robust to label noise. This is important in practice
especially when we are using deep neural networks which tend to
overfit noise. These results have been published in NeurIPS, ECCV,
ICML and ICLR.

Fall 2026, Wednesdays 2 to 3:20 pm, NCS 220 and Zoom link to be announced soon.

The seminar will be jointly taught by Prof. Dimitris Samaras (samaras@cs.stonybrook.edu).

The overall purpose of this seminar is to bring together people with interests in Computer Vision theory and techniques and to examine current research issues. This course will be appropriate for people who already took a Computer Vision graduate course or already had research experience in Computer Vision.

To enroll in this course, you must either: (1) be in the Ph.D. program or (2) receive permission from the instructors.

Each seminar will consist of multiple short talks (around 15 minutes) by multiple students. Students can register for 1 credit for CSE656. Registered students must attend and present a minimum of 2 talks. Registered students must attend in person. Up to 3 absences will be excused. Everyone else is welcome!

Abstract: Generative image models trained on massive datasets encode the statistics of our visual world. An off-the-shelf diffusion or flow-matching model can therefore serve as a general-purpose image prior, providing information about images across a wide range of tasks. Harnessing these capabilities, however, requires combining the model at inference time with additional signals or constraints unknown during training. This thesis will develop training-free methods for inference under a generative denoising prior. I first show how sampling from a trained denoiser can be formulated as an optimization problem, which combined with a constraint, can solve tasks ranging from conditional generation and weakly supervised segmentation to combinatorial optimization. I then demonstrate how the inherent properties of denoisers can accelerate inference under such constraints. Replacing gradient descent with an inexact Newton update, based on the symmetry of the denoiser's Jacobian, substantially reduces inference costs without tradeoffs. I also explore a middle ground between model adaptation and fully training-free inference by using the denoiser's robust internal representations to learn constraints from limited labeled data. These methods are applied to gigapixel image domains such as digital histopathology and remote sensing, where generative models can only be trained at a patch scale. To synthesize arbitrarily large images at resolutions unseen during training, I introduce inference-time algorithms that enforce consistency across spatially overlapping patches and image scales. Finally, I propose repurposing inference-time algorithms from sampling tools, to mechanisms for understanding the prior learned by a denoising model. Extending the previously developed techniques, I analyze the Jacobians of generative denoisers, where preliminary results indicate that Jacobian spectra correlates with generative quality. This motivates a Jacobian-spectrum regularization as a way to improve model performance using insights derived from inference-time algorithms.

Speaker: Alexandros Graikos

Location: NCS 220

Abstract: Modern language agents often need to solve tasks requiring long-horizon, multi-turn interactions, where they retrieve external information, adapt to observations, and answer interdependent queries. Yet, most LLM systems rely on full-context prompting, appending all past turns regardless of their relevance. This leads to un-bounded memory growth, increased computational costs, and degraded reasoning performance on out-of-distribution input lengths due to LLM forgetting the context. We introduce MEM1, an end-to-end reinforcement learning framework that enables agents to operate with constant context size when solving long multi-turn tasks. At each turn, MEM1 updates a compact shared internal state that jointly supports memory consolidation and reasoning. Leveraging reinforcement learning (RL) and rollout trajectory truncation, we train a MEM1 agent to develop internal states that integrate prior memory with new observations from the environment while strategically discarding irrelevant or redundant information. Experiments across three domains, including internal retrieval QA, open-domain web QA, and multi-turn web shopping, show that MEM1-7B improves performance by 3.5x while reducing memory usage by 3.7x compared to Qwen2.5-14B-Instruct on an augmented multi-hop QA dataset with 16 objectives in each task, and generalizes beyond the training horizon. Our results demonstrate the promise of reasoning-driven memory consolidation as a scalable alternative to existing solutions for training long-horizon task-solving agents that involve multiple interactions, where both efficiency and performance are optimized.

Speaker: Yiyang Feng

Location: CS2311

University Libraries Present: Analyzing quantitative data can feel overwhelming without the right tools. In this workshop, SBU Libraries' Data Literacies Lead, Ahmad Pratama will show you how to master the basics of exploratory data analysis for quantitative data using Python. This workshop covers several techniques to help you uncover patterns and insights in your datasets.

Online RSVP via link: https://stonybrook.zoom.us/meeting/register/vEPycmDrQoGjFqkmsYHgxw

Topic: AI Seminar: Owen Rambow
Time: Mar 17, 2021 10:00 AM Eastern Time (US and Canada)
Join Zoom Meeting

https://stonybrook.zoom.us/j/93614644178?pwd=MzJtVDJYYmU5T1dtMzJiUFMxb0x4dz09
Meeting ID: 936 1464 4178.    Passcode: 965936






Natural Language Understanding and Semantic Parsing

(Partly joint work with former colleagues at Elemental Cognition)

Semantic parsing refers to the task of determining the propositional content of language: who did what to whom.  It is part of the larger task of natural language understanding (NLU).  I will start out by discussing what full NLU means, and argue that we are still far away, as a field, from solving full NLU, or even from knowing how to evaluate it.

In the second part of the talk, I will situate semantic parsing in the context of several other NLU subtasks.  Typically, the target representation of semantic parsing uses an ontology (such as PropBank or FrameNet).  Semantic parsing includes the subtasks of word sense disambiguation, argument detection, and argument role labeling.  I will discuss choices among possible target ontologies.  I will justify why we created a new ontology, Hector, based on FrameNet and the lexical resource NOAD, and explain some of its characteristics.

In the third part of the talk, I will present experiments we performed using transformer models.  We obtain best results using a two-phase model, in which we first choose the frame, and then, given the frame, choose the arguments.  We encode the problem for both tasks using indices in the sentence.  While we develop the parser for our new ontology Hector, this approach also beats the state of the art for FrameNet and PropBank parsing.Biography:  I am a professor in the Department of Linguistics at Stony Brook University with a joint appointment in IACS.

Until recently, I was a research scientist at Elemental Cognition. Elemental Cognition is working on deep natural language understanding.

I got my PhD with Aravind Joshi at the University of Pennsylvania in 1994. I have worked at CoGenTex, and at AT&T Labs -- Research, and for many years I was a research scientist at Columbia University in the Center for Computational Learning Systems.

Join C4E for an interactive workshop with Dr. Raju Rajan, founder of ReWild Long Island, exploring how artificial intelligence can support community-led land stewardship.

ReWild Long Island advances hands-on ecological restoration across Long Island through native planting, organic land care, water protection, composting, and waste reduction.

Date & Time: Wednesday, November 4 | 12:30 PM - 1:50 PM

Location: Wang Center, Room 301

Three-Part Agenda Breakdown:

1. The Vision: ReWild's AI Tool
2. The Community: Shaping the Tool Together
   
3. The Standards: Applying Responsible AI Frameworks

Speaker: Raju Rajan is a technologist with a strong communitarian ethic. Since obtaining a PhD in Communication Networks at the University of Wisconsin, Madison, he has enjoyed a long career as researcher and consultant. His entrepreneurial acumen has led him to found two IT start-ups as well as a number of community groups aimed at organizing people for systemic change.
Raju is one of the founders of ReWild Long Island and currently serves as Board President. He is works with Long Island Together, a progressive collective that organizes around education, immigration and social change.
An avid gardener and amateur tree-enthusiast, Raju also owns and manages Sharadavanam, a small agroforest farm in Southern India dedicated to rewilding and permaculture.
Raju resides in Port Washington, NY with wife Sonia Arora, an educator & community organizer, and son Kabeera Singh, in college.

Please register using this form to join the discussion.