Conference Overview
Machine learning for scientific imaging is a rapidly growing area of research used to characterize physical, material, chemical, and biological processes in both large and small scale scientific experiments. Physics inspired machine learning differs from more general machine learning research in that it emphasizes quantitative reproducibility and the incorporation of physical models. ML methods used for scientific imaging typically incorporate physics-based imaging processes or physics-based models of the underlying data. These models can be based on partial differential equations (PDEs), integral equations, symmetries or other regularity conditions in two or more dimensions. Physics aware models enhance the ability of the ML methods to generalize and robustly operate in the presence of modeling error, incomplete data, and measurement uncertainty. Contributions to the conference are solicited on topics ranging from fundamental theoretical advances to detailed implementations and novel applications for scientific discovery.
2024 Conference Topics
Algorithms and methodologies
- Machine learning-inspired scientific imaging system design
- Learning-based modeling and algorithms for scientific imaging
- Novel machine learning theory for scientific imaging
- Physics-inspired machine learning theory and applications
- Machine learning-inspired new physics
Key problem areas
- Tomography (transmission and emission)
- NMR, MRI
- Diffraction theory
- Synchrotron X-ray imaging
- XFEL
- Crystallography
- Synthetic aperture radar
- Electron microscopy
- Structural biology
- Phase retrieval
- Phase contrast imaging
- Fourier ptychography
- Astronomical imaging
- Physics based machine learning of dynamic processes
2024 Special Sessions
The Intersection of Computational Imaging and Materials Science
Session Organizing Chairs:
Jeff P. Simmons, Air Force Research Laboratory (US)
Yuejie Chi, Carnegie Mellon University (US)
B. Reeja Jayan, Carnegie Mellon University (US)
Advances in machine learning echo through the halls of traditional fields, changing the very way we view these subjects. Yet the enigmatic complexity of materials science sets it apart, demanding multi-scale observations of hierarchical structure with a myriad of microscopic techniques. Yet, within the shadows, the spirit of machine learning lurks, and computational imaging techniques, such as advanced microscopy, tomography, and spectroscopy, are seen in a new light, allowing unprecedented advances, transcending the boundaries of the unknown. Efficient extraction of complex structural and chemical information from materials at various scales and fusion of high-dimensional imaging datasets enables the automatic identification of key material features, leading to accelerated material discovery and characterization. An apparition of possibilities emerges enabling the prediction of material behavior, degradation, and performance under different conditions for designing innovative materials for applications in energy, electronics, medicine, and beyond. The committee invites contributions exploring the advancements in data acquisition, image reconstruction, and feature extraction techniques. Additionally, the organizing chairs encourage presentations on applying deep neural networks, generative models, graph theory and topology, optimization methods, inverse methods, physics integration with learning, and other emerging techniques toward enhancing materials analysis and accelerating the translation of research findings into real-world applications.
This is a joint special session with Computational Imaging XXII.
Large Language Models for Scientific Discovery
2024 Committee
Conference Chairs
Marc Louis Klasky, Los Alamos National Laboratory (United States)
Jong Chul Ye, Korea Advanced Institute of Science & Technology (KAIST) (Republic of Korea)
Program Committee
Raja Giryes, Tel Aviv University (Israel)
Ulugbek Kamilov, Washington University in St. Louis (United States)
Christopher Vincent Rackauckas, Massachusetts Institute of Technology (United States)
Daniel M. Tartakovsky, Stanford University (United States)
Lei Tian, Boston University (United States)
Nathaniel Trask, Sandia National Laboratories (United States)