We invite researchers and practitioners working at the intersection of AI, Physics, and Geophysics to submit their work to the "AI in Physical Sciences: Toward a Unified Approach for Physics and Geophysics".
Format: Abstracts between 150 and 500 words
Multiple submissions are allowed per participant
Language: English
All submissions will be peer-reviewed by the program committee based on relevance, clarity, and alignment with the workshop themes.
Selected abstracts will be presented during:
4 Oral Sessions (the best selected one for each main topic)
Poster Session for all other accepted abstracts. The poster format is 150x90 Portrait.
Call for Abstracts Opens: May 15, 2025
Submission Deadline: June 30, 2025 extended to July 15, 2025
Notification of Acceptance: July 30, 2025
We are particularly interested in contributions addressing:
We are particularly interested in contributions addressing the following themes:
Physics-based Neural Networks for solving PDEs in seismology, electromagnetism, and fluid dynamics
Neural Operators (e.g., Fourier Neural Operators, DeepONets) for spatiotemporal systems
Interpretable Machine Learning for transparent scientific modeling
Uncertainty Quantification using Bayesian deep learning and ensembles
Synergy of Data-driven and Physics-driven AI approaches
Multi-scale and Multi-resolution Learning across physical phenomena
Hybrid AI-Physical Models, integrating AI with FEM, CFD, etc.
Anomaly Detection in Scientific Data, e.g., rare events in physics or geophysics
Self-Supervised and Unsupervised Learning for sparse and noisy scientific data
AI for Inverse Problems in tomography, remote sensing, gravitational wave detection
Causal Inference in AI for better generalization in physical modeling
AI + HPC for Large-Scale Scientific Simulations
Foundation Models tailored for physics and geoscience applications
Graph Neural Networks in scientific modeling (e.g., tectonics, molecular structures)
Time-Series Forecasting in astrophysics, climate science, and space weather
Generative AI (e.g., GANs, diffusion models) for simulations and data augmentation
Open Science and Reproducibility in AI for scientific research
Present your work to a targeted interdisciplinary audience
Network with experts from top institutions in AI, physics, and geophysics
Be considered for an oral presentation in one of the four main sessions
Gain visibility at a unique, high-impact scientific event
Participation is free of charge, with priority given to submitting authors. We anticipate hosting a maximum of 80 attendees.
Submissions will open on May 15, 2025. A dedicated submission portal will be made available here. Stay tuned!
For any questions regarding the submission process, feel free to contact the organizing committee at: [Insert contact email]