Le Hou Dissertation Defense: Deep Learning for Digital Histopathology across Multiple Scales
ABSTRACT: Histopathology is the study of tissue changes caused by diseases such as cancer. It plays a crucial role in disease diagnosis, survival analysis and development of new treatments. Using computer vision techniques, I focus on multiple tasks for automated analysis in digital histopathology images, which are challenging because histopathology images are heterogeneous and complex, due to the large variation of hundreds of cancer types in gigapixel resolution. In this thesis, I show how histopathology image analysis tasks can be viewed in three scales: Whole Slide Image (WSI)-level, patch-level and cellular-level, and present my contributions in each resolution level.
BIO: WSI-level analysis such as classifying WSIs into cancer types is challenging, because conventional classification methods such as off-the-shelf deep learning models cannot be applied directly on gigapixel WSIs due to computational limitations. I contribute a patch-based deep learning method that classifies gigapixel WSIs into cancer types and subtypes with close-to-human performance. This method is useful for computer-aided diagnosis. At patch-level, I contribute a novel method for histopathology image patch classification. On the task of identifying Tumor Infiltrating Lymphocyte (TIL) regions, the prediction result of this method correlates to the survival rate of patients. At cellular-level, I contribute novel methods for nucleus classification and roundness regression, which are interpretable features for histopathology studies. With this method, I generated a large-scale dataset of segmented nuclei, in WSIs from a large publicly available digital histopathology image dataset, to help advance histopathology research.
ABSTRACT: Histopathology is the study of tissue changes caused by diseases such as cancer. It plays a crucial role in disease diagnosis, survival analysis and development of new treatments. Using computer vision techniques, I focus on multiple tasks for automated analysis in digital histopathology images, which are challenging because histopathology images are heterogeneous and complex, due to the large variation of hundreds of cancer types in gigapixel resolution. In this thesis, I show how histopathology image analysis tasks can be viewed in three scales: Whole Slide Image (WSI)-level, patch-level and cellular-level, and present my contributions in each resolution level.
BIO: WSI-level analysis such as classifying WSIs into cancer types is challenging, because conventional classification methods such as off-the-shelf deep learning models cannot be applied directly on gigapixel WSIs due to computational limitations. I contribute a patch-based deep learning method that classifies gigapixel WSIs into cancer types and subtypes with close-to-human performance. This method is useful for computer-aided diagnosis. At patch-level, I contribute a novel method for histopathology image patch classification. On the task of identifying Tumor Infiltrating Lymphocyte (TIL) regions, the prediction result of this method correlates to the survival rate of patients. At cellular-level, I contribute novel methods for nucleus classification and roundness regression, which are interpretable features for histopathology studies. With this method, I generated a large-scale dataset of segmented nuclei, in WSIs from a large publicly available digital histopathology image dataset, to help advance histopathology research.
Simons Laufer Mathematical Sciences Institute presents...
In 2023, Tudor Achim co-founded Harmonic with Vlad Tenev to build the world's most advanced reasoning engine. Combining formal verification with informal reasoning, Harmonic's formal reasoning model, Aristotle, achieved gold-medal-equivalent performance on the 2025 International Mathematical Olympiad problems. Aristotle integrates three main components: a Lean proof search system, an informal reasoning system that generates and formalizes lemmas, and a dedicated geometry solver.
Achim is also the Co-Founder and former CTO of Helm.ai. He holds a B.S. in Computer Science from Carnegie Mellon University and was a PhD Candidate in Computer Science at Stanford University.
Register here: https://slmath.us10.list-manage.com/track/click?u=d58ee2e82c69809ff037f56b2&id=f07a675f6f&e=f1b6ba91e6
Abstract: Machine learning (ML) systems fueled by neural networks have entered our daily lives and led to scientific breakthroughs, but many open questions remain. After a nod toward the question of rigor with ML and recent progress, I'll turn to the theory of neural networks. I will argue that understanding neural networks inevitably leads to ideas from field theory (FT), which was already realized in the simplest case in the 1990s, and I will review some essential FT-for-NN results. I will then propose that the connection might be more general, an NN-FT correspondence of sorts, with neural networks providing a way to define a field theory. I'll end with comments on known results including the origin of interactions and various symmetries, but I will also list some open questions. The apparent non-sequitur in the title will be used as a rhetorical device to explore where we are and where we'd like to go.
https://scgp.stonybrook.edu/calendar/full-calendar
https://scgp.stonybrook.edu/calendar/full-calendar
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
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
How to Succeed in Language Design Without Really Trying presented by Professor Brian Kernighan
ABSTRACT: Why do some languages succeed while others fall by the wayside? I've helped create nearly a dozen languages (mostly small) over the years; a handful are still in widespread use, while others have languished or simply disappeared. I've also been present at the creation of several other languages, including some really major ones. In this talk I'll give my humble, but correct, opinion on factors that affect success and failure, and try to offer some insight into what to do if you're trying to design a new language yourself, and why that might be a good thing.
BIO: Brian Kernighan received a PhD in electrical engineering from Princeton in 1969. He joined the Computer Science department at Princeton in 2000, after many years at Bell Labs. He is a co-creator of several programming languages, including AWK and AMPL, and of a number of tools for document preparation. He is the co-author of a dozen books and some technical papers, and holds 5 patents.
He is a member of the National Academy of Engineering and of the American Academy of Arts and Sciences. His research areas include programming languages, tools and interfaces that make computers easier to use, often for non-specialist users. He has also written two books on technology for
non-technical audiences: Understanding the Digital World in 2017 and Millions, Billions, Zillions: Defending Yourself in a World of Too Many Numbers, published in 2018. His most recent book, Unix: A History and a Memoir, was published in October 2019.
ABSTRACT: Why do some languages succeed while others fall by the wayside? I've helped create nearly a dozen languages (mostly small) over the years; a handful are still in widespread use, while others have languished or simply disappeared. I've also been present at the creation of several other languages, including some really major ones. In this talk I'll give my humble, but correct, opinion on factors that affect success and failure, and try to offer some insight into what to do if you're trying to design a new language yourself, and why that might be a good thing.
BIO: Brian Kernighan received a PhD in electrical engineering from Princeton in 1969. He joined the Computer Science department at Princeton in 2000, after many years at Bell Labs. He is a co-creator of several programming languages, including AWK and AMPL, and of a number of tools for document preparation. He is the co-author of a dozen books and some technical papers, and holds 5 patents.
He is a member of the National Academy of Engineering and of the American Academy of Arts and Sciences. His research areas include programming languages, tools and interfaces that make computers easier to use, often for non-specialist users. He has also written two books on technology for
non-technical audiences: Understanding the Digital World in 2017 and Millions, Billions, Zillions: Defending Yourself in a World of Too Many Numbers, published in 2018. His most recent book, Unix: A History and a Memoir, was published in October 2019.
Abstract: Many scientific and engineering challenges, such as the design of materials or molecules or the control of experimental systems, rely on the existence of fast predictive models that can evaluate potential designs or control policies. Traditionally this has been accomplished through numerical simulation; more recently data-driven machine learning methods have been applied. However, both approaches leave gaps: physical modeling can be accurate and extrapolates well to previously-unstudied conditions, but it is often computationally expensive and relies on physics approximations that may not be valid. Machine learning can generalize from massive amounts of real-world or simulation data, but suffers from physical grounding and extrapolation into new regimes, as well as in settings where large data sets do not exist.
In this talk I explore an intermediate regime, which is hybrid reduced order models: fast simplified physics approximations where some of the unknown or approximated equations are replaced with data-driven machine learning components. Examples include coarse-grained models where the full macroscopic equations cannot be derived from first-principles microscopic equations, multiscale models with unknown closure terms or sub-grid parameterization schemes, and low-order or latent dynamical systems that learn governing equations on a low-dimensional reduced state space. I discuss how such reduced systems can be identified from very limited data, much less than is often needed in traditional machine learning but at much lower time-to-solution than traditional numerical modeling. This facilitates not only system design and control but also uncertainty quantification approaches that search the space of possible equations for predictive models that can explain the data. I will focus on an example from materials science concerning the design of self-assembling block copolymer nanomaterials.
Speaker: Dr. Nathan Urban, Applied Mathematics Department, Brookhaven National Laboratory
Location: Laufer 101
Zoom: https://stonybrook.zoom.us/j/96090260834?pwd=mw8QTHbMOw9oeU9hazZeoq8bN4VIfH.1
Meeting ID: 960 9026 0834 Passcode: 374969
In this talk I explore an intermediate regime, which is hybrid reduced order models: fast simplified physics approximations where some of the unknown or approximated equations are replaced with data-driven machine learning components. Examples include coarse-grained models where the full macroscopic equations cannot be derived from first-principles microscopic equations, multiscale models with unknown closure terms or sub-grid parameterization schemes, and low-order or latent dynamical systems that learn governing equations on a low-dimensional reduced state space. I discuss how such reduced systems can be identified from very limited data, much less than is often needed in traditional machine learning but at much lower time-to-solution than traditional numerical modeling. This facilitates not only system design and control but also uncertainty quantification approaches that search the space of possible equations for predictive models that can explain the data. I will focus on an example from materials science concerning the design of self-assembling block copolymer nanomaterials.
Speaker: Dr. Nathan Urban, Applied Mathematics Department, Brookhaven National Laboratory
Location: Laufer 101
Zoom: https://stonybrook.zoom.us/j/96090260834?pwd=mw8QTHbMOw9oeU9hazZeoq8bN4VIfH.1
Meeting ID: 960 9026 0834 Passcode: 374969
Are you interested in understanding the challenges that lie ahead as Artificial Intelligence (AI) systems become increasingly autonomous, dynamically acquire information, and adapt behaviors?
Join us for an exciting afternoon of talks by visionaries and leaders from industry, government, and academia as we kickoff a three-part Trusted AI Challenge Series designed to Build the Vision - Formalize Challenges - Advance the Art of next generation of AI systems.
The Air Force Research Laboratory Information Directorate, The State University of New York, Innovare Advancement Center, NYSTEC, and Griffiss Institute invite you to join us for this half-day virtual event!
WHEN: Wednesday, October 14, 2020, 12:00 PM - 4:00 PM EDT
Hosted by Innovare Advancement Center, this webinar is the first of a three-part series designed to cultivate, define and fund creative solutions to a set of challenge problems in trustworthy AI with a particular focus on dynamic, autonomous systems that learn and adapt behaviors.
Keynote speakers include Dr. David Goldstein of Space X; Dr. Scott Hubbard of Stanford University; Dr. Pramod Khargonekar of UC Irvine, and more!
This event is designed for academic and government researchers, university students, and small businesses.
Would you like to understand some of the most formidable technical challenges in future autonomous systems? Would you like to sponsor some of the brightest minds in AI to work on problems of interest to you? Would you like to learn more about AI in real systems?
If so, Save the Date! Wednesday, October 14, 2020, 12:00 PM - 4:00 PM EDT.
Please see additional information on the three-part series here. Registration details to follow!
Stay tuned: https://www.innovare.org/news-events
Join us for an exciting afternoon of talks by visionaries and leaders from industry, government, and academia as we kickoff a three-part Trusted AI Challenge Series designed to Build the Vision - Formalize Challenges - Advance the Art of next generation of AI systems.
The Air Force Research Laboratory Information Directorate, The State University of New York, Innovare Advancement Center, NYSTEC, and Griffiss Institute invite you to join us for this half-day virtual event!
WHEN: Wednesday, October 14, 2020, 12:00 PM - 4:00 PM EDT
Hosted by Innovare Advancement Center, this webinar is the first of a three-part series designed to cultivate, define and fund creative solutions to a set of challenge problems in trustworthy AI with a particular focus on dynamic, autonomous systems that learn and adapt behaviors.
Keynote speakers include Dr. David Goldstein of Space X; Dr. Scott Hubbard of Stanford University; Dr. Pramod Khargonekar of UC Irvine, and more!
This event is designed for academic and government researchers, university students, and small businesses.
Would you like to understand some of the most formidable technical challenges in future autonomous systems? Would you like to sponsor some of the brightest minds in AI to work on problems of interest to you? Would you like to learn more about AI in real systems?
If so, Save the Date! Wednesday, October 14, 2020, 12:00 PM - 4:00 PM EDT.
Please see additional information on the three-part series here. Registration details to follow!
Stay tuned: https://www.innovare.org/news-
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 PhD program or (2) receive permission from the instructors.
Each seminar will consist of multiple short talks (around 10 minutes) by multiple people. Students can register for 1 credit for CSE 656. Registered students must attend and present a minimum of 2 or 3 talks. Everyone else is welcome to attend. Fill in https://forms.gle/pCVXovgfMfQwGqG38 to subscribe to our mailing list for further announcement.
Each seminar will consist of multiple short talks (around 10 minutes) by multiple people. Students can register for 1 credit for CSE 656. Registered students must attend and present a minimum of 2 or 3 talks. Everyone else is welcome to attend. Fill in https://forms.gle/pCVXovgfMfQwGqG38 to subscribe to our mailing list for further announcement.