Emerging Flexible Designs for Geospatial Multimodal Foundation Models

Foundation models are rapidly transforming Earth observation by enabling scalable pretraining across diverse unlabeled geospatial modalities. However, their architectural diversity ranging from encoder-only to encoder-decoder and masked autoencoding paradigms makes it challenging to assess performance trade offs in a consistent manner. In this work, we present an apples-to-apples comparison of leading FM architectures designed for geospatial multimodal reasoning, with a particular focus on flexibility across varied spectral band configurations. We standardize pretraining using identical self supervised learning objectives and training datasets, and evaluate all models under consistent parameterization on the GEOBench benchmark across classification and segmentation tasks. Our results offer new insights into the design trade-offs between model flexibility, modality alignment, and downstream task performance. By highlighting architectural strengths and limitations under controlled conditions, this study provides practical guidance for building next generation geospatial foundation models capable of robust multimodal reasoning.

GeoGNN: Time Series Geo-Localization using Two-Tower Graph Neural Networks

This paper investigates a novel concept of time series geolocalization, where the goal is to infer the geographic origin of each raw time series. Successful geolocalization can provide spatial context to time series, enabling downstream location-aware applications. We formalize the problem, adapt core ideas from image geolocalization to establish strong baselines, and propose GeoGNN, a two-tower architecture. During training, GeoGNN's spatial tower learns embeddings of geographic cell candidates by leveraging the geographic adjacency graph, while the temporal tower extracts informative representations from time series. During inference, each temporal representation is matched against candidate geographic embeddings using dot-product similarity, combined with an auxiliary classification head, to predict the time series' associated geographic origin. Experiments on large-scale, countrywide electricity-consumption datasets demonstrate that GeoGNN achieves the best performance across datasets and enhances both fine- and coarse-grained geolocalization accuracy by ~27% on average.

SimLBR: Learning to Detect Fake Images by Learning to Detect Real Images

The rapid advancement of generative models has made the detection of AI-generated images a critical challenge for both research and society. Recent works have shown that most state-of-the-art fake image detection methods overfit to their training data and catastrophically fail when evaluated on curated hard test sets with strong distribution shifts. In this work, we argue that it is more principled to learn a tight decision boundary around the real image distribution and treat the fake category as a sink class. To this end, we propose SimLBR, a simple and efficient framework for fake image detection using Latent Blending Regularization (LBR). Our method significantly improves cross-generator generalization, achieving up to +24.85\% accuracy and +69.62\% recall on the challenging Chameleon benchmark. SimLBR is also highly efficient, training orders of magnitude faster than existing approaches. Furthermore, we emphasize the need for reliability-oriented evaluation in fake image detection, introducing risk-adjusted metrics and worst-case estimates to better assess model robustness. All code and models will be released on HuggingFace and GitHub.

OReole-FM: successes and challenges toward billion-parameter foundation models for high-resolution satellite imagery

While the pretraining of Foundation Models (FMs) for remote sensing (RS) imagery is on the rise, models remain restricted to a few hundred million parameters. Scaling models to billions of parameters has been shown to yield unprecedented benefits including emergent abilities, but requires data scaling and computing resources typically not available outside industry R&D labs. In this work, we pair high-performance computing resources including Frontier supercomputer, America's first exascale system, and high-resolution optical RS data to pretrain billion-scale FMs. Our study assesses performance of different pretrained variants of vision Transformers across image classification, semantic segmentation and object detection benchmarks, which highlight the importance of data scaling for effective model scaling. Moreover, we discuss construction of a novel TIU pretraining dataset, model initialization, with data and pretrained models intended for public release. By discussing technical challenges and details often lacking in the related literature, this work is intended to offer best practices to the geospatial community toward efficient training and benchmarking of larger FMs.

Pretraining Billion-scale Geospatial Foundational Models on Frontier

As AI workloads increase in scope, generalization capability becomes challenging for small task-specific models and their demand for large amounts of labeled training samples increases. On the contrary, Foundation Models (FMs) are trained with internet-scale unlabeled data via self-supervised learning and have been shown to adapt to various tasks with minimal fine-tuning. Although large FMs have demonstrated significant impact in natural language processing and computer vision, efforts toward FMs for geospatial applications have been restricted to smaller size models, as pretraining larger models requires very large computing resources equipped with state-of-the-art hardware accelerators. Current satellite constellations collect 100+TBs of data a day, resulting in images that are billions of pixels and multimodal in nature. Such geospatial data poses unique challenges opening up new opportunities to develop FMs. We investigate billion scale FMs and HPC training profiles for geospatial applications by pretraining on publicly available data. We studied from end-to-end the performance and impact in the solution by scaling the model size. Our larger 3B parameter size model achieves up to 30% improvement in top1 scene classification accuracy when comparing a 100M parameter model. Moreover, we detail performance experiments on the Frontier supercomputer, America's first exascale system, where we study different model and data parallel approaches using PyTorch's Fully Sharded Data Parallel library. Specifically, we study variants of the Vision Transformer architecture (ViT), conducting performance analysis for ViT models with size up to 15B parameters. By discussing throughput and performance bottlenecks under different parallelism configurations, we offer insights on how to leverage such leadership-class HPC resources when developing large models for geospatial imagery applications.

DeepSpeed4Science Initiative: Enabling Large-Scale Scientific Discovery through Sophisticated AI System Technologies

In the upcoming decade, deep learning may revolutionize the natural sciences, enhancing our capacity to model and predict natural occurrences. This could herald a new era of scientific exploration, bringing significant advancements across sectors from drug development to renewable energy. To answer this call, we present DeepSpeed4Science initiative (deepspeed4science.ai) which aims to build unique capabilities through AI system technology innovations to help domain experts to unlock today's biggest science mysteries. By leveraging DeepSpeed's current technology pillars (training, inference and compression) as base technology enablers, DeepSpeed4Science will create a new set of AI system technologies tailored for accelerating scientific discoveries by addressing their unique complexity beyond the common technical approaches used for accelerating generic large language models (LLMs). In this paper, we showcase the early progress we made with DeepSpeed4Science in addressing two of the critical system challenges in structural biology research.

Proceedings of KDD 2021 Workshop on Data-driven Humanitarian Mapping: Harnessing Human-Machine Intelligence for High-Stake Public Policy and Resilience Planning

Humanitarian challenges, including natural disasters, food insecurity, climate change, racial and gender violence, environmental crises, the COVID-19 coronavirus pandemic, human rights violations, and forced displacements, disproportionately impact vulnerable communities worldwide. According to UN OCHA, 235 million people will require humanitarian assistance in 2021. Despite these growing perils, there remains a notable paucity of data science research to scientifically inform equitable public policy decisions for improving the livelihood of at-risk populations. Scattered data science efforts exist to address these challenges, but they remain isolated from practice and prone to algorithmic harms concerning lack of privacy, fairness, interpretability, accountability, transparency, and ethics. Biases in data-driven methods carry the risk of amplifying inequalities in high-stakes policy decisions that impact the livelihood of millions of people. Consequently, proclaimed benefits of data-driven innovations remain inaccessible to policymakers, practitioners, and marginalized communities at the core of humanitarian actions and global development. To help fill this gap, we propose the Data-driven Humanitarian Mapping Research Program, which focuses on developing novel data science methodologies that harness human-machine intelligence for high-stakes public policy and resilience planning. The proceedings of the 2nd Data-driven Humanitarian Mapping workshop at the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining. August 15th, 2021

Proceedings of KDD 2020 Workshop on Data-driven Humanitarian Mapping: Harnessing Human-Machine Intelligence for High-Stake Public Policy and Resilience Planning

Humanitarian challenges, including natural disasters, food insecurity, climate change, racial and gender violence, environmental crises, the COVID-19 coronavirus pandemic, human rights violations, and forced displacements, disproportionately impact vulnerable communities worldwide. According to UN OCHA, 235 million people will require humanitarian assistance in 2021 . Despite these growing perils, there remains a notable paucity of data science research to scientifically inform equitable public policy decisions for improving the livelihood of at-risk populations. Scattered data science efforts exist to address these challenges, but they remain isolated from practice and prone to algorithmic harms concerning lack of privacy, fairness, interpretability, accountability, transparency, and ethics. Biases in data-driven methods carry the risk of amplifying inequalities in high-stakes policy decisions that impact the livelihood of millions of people. Consequently, proclaimed benefits of data-driven innovations remain inaccessible to policymakers, practitioners, and marginalized communities at the core of humanitarian actions and global development. To help fill this gap, we propose the Data-driven Humanitarian Mapping Research Program, which focuses on developing novel data science methodologies that harness human-machine intelligence for high-stakes public policy and resilience planning. The proceedings of the 1st Data-driven Humanitarian Mapping workshop at the 26th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, August 24th, 2020.

Computer-aided abnormality detection in chest radiographs in a clinical setting via domain-adaptation

Deep learning (DL) models are being deployed at medical centers to aid radiologists for diagnosis of lung conditions from chest radiographs. Such models are often trained on a large volume of publicly available labeled radiographs. These pre-trained DL models' ability to generalize in clinical settings is poor because of the changes in data distributions between publicly available and privately held radiographs. In chest radiographs, the heterogeneity in distributions arises from the diverse conditions in X-ray equipment and their configurations used for generating the images. In the machine learning community, the challenges posed by the heterogeneity in the data generation source is known as domain shift, which is a mode shift in the generative model. In this work, we introduce a domain-shift detection and removal method to overcome this problem. Our experimental results show the proposed method's effectiveness in deploying a pre-trained DL model for abnormality detection in chest radiographs in a clinical setting.

Apache Spark Accelerated Deep Learning Inference for Large Scale Satellite Image Analytics

The shear volumes of data generated from earth observation and remote sensing technologies continue to make major impact; leaping key geospatial applications into the dual data and compute intensive era. As a consequence, this rapid advancement poses new computational and data processing challenges. We implement a novel remote sensing data flow (RESFlow) for advanced machine learning and computing with massive amounts of remotely sensed imagery. The core contribution is partitioning massive amount of data based on the spectral and semantic characteristics for distributed imagery analysis. RESFlow takes advantage of both a unified analytics engine for large-scale data processing and the availability of modern computing hardware to harness the acceleration of deep learning inference on expansive remote sensing imagery. The framework incorporates a strategy to optimize resource utilization across multiple executors assigned to a single worker. We showcase its deployment across computationally and data-intensive on pixel-level labeling workloads. The pipeline invokes deep learning inference at three stages; during deep feature extraction, deep metric mapping, and deep semantic segmentation. The tasks impose compute intensive and GPU resource sharing challenges motivating for a parallelized pipeline for all execution steps. By taking advantage of Apache Spark, Nvidia DGX1, and DGX2 computing platforms, we demonstrate unprecedented compute speed-ups for deep learning inference on pixel labeling workloads; processing 21,028~Terrabytes of imagery data and delivering an output maps at area rate of 5.245sq.km/sec, amounting to 453,168 sq.km/day - reducing a 28 day workload to 21~hours.