$\textit{One Stone Three Birds:}$ Three-Dimensional Implicit Neural Network for Compression and Continuous Representation of Multi-Altitude Climate Data

ICLR 2025 Conference Submission13370 Authors

Keywords: Multi-modal representation learning, continuous super-resolution, dimensionality reduction, cross-modal prediction, scientific data compression

Abstract:

Wind energy stands out as a promising clean and renewable energy alternative, not only for its potential to combat global warming but also for its capacity to meet the ever-growing demand for energy. However, analysis of wind data to fully harness the benefits of wind energy demands tackling several related challenges:
(1) Current data resolution is inadequate for capturing the detailed information needed across diverse climatic conditions;
(2) Efficient management and storage of real-time measurements are currently lacking;
(3) Extrapolating wind data across spatial specifications enables analysis at costly-to-measure, unobserved points is necessary.
In response to these challenges, we introduce a modality-agnostic learning framework utilizing implicit neural networks. Our model effectively compresses a large volume of climate data into a manageable latent codec. It also learns underlying continuous climate patterns, enabling reconstruction at any scale and supporting modality transfer and fusion. Extensive experimental results show consistent performance improvements over existing baselines.

Primary Area: applications to physical sciences (physics, chemistry, biology, etc.)

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.

Reciprocal Reviewing: I understand the reciprocal reviewing requirement as described on https://iclr.cc/Conferences/2025/CallForPapers. If none of the authors are registered as a reviewer, it may result in a desk rejection at the discretion of the program chairs. To request an exception, please complete this form at https://forms.gle/Huojr6VjkFxiQsUp6.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 13370