February 28th, 2025 – Tucson, AZ
The 2nd Workshop on Computer Vision for Earth Observation (CV4EO) Applications is conceived as a platform to foster application-oriented discussions between the computer vision community and experts from geoscience domains, remote sensing data providers, governmental agencies, and other organizations utilizing computer vision-enabled EO data analysis for decision-making in disaster response, national security, environmental protection, and other application areas.
The workshop aims to achieve the following goals:
Promote Multidisciplinary & Cross-Sectoral Interaction toward impactful applications: The workshop will expose members of the CV community to the exciting and impactful applications within the EO domain, discussing real-world applications such as decision-making in disaster response, national security, and environmental protection. CV for EO is a multidisciplinary space with diverse data sources, application domains, and involved disciplines. Addressing its open challenges and opportunities requires interaction ranging from stakeholders, problem owners, to experts on the research and development of data analysis tools. This workshop will foster knowledge exchange and collaborative efforts across multiple disciplines and sectors (e.g., government agencies, data providers, industry, national laboratories, academic researchers)
Address Challenges in Multimodal Data: Data from existing remote sensing modalities are heterogeneous and complementary in many aspects. Passive imagery sources vary significantly in number of channels (e.g., multispectral vs. hyperspectral data), while active imagery such as Synthetic Aperture Radar (SAR) include both amplitude and phase components. Data representation formats also vary: geospatial data is often represented in vector format, and their rasterization for consumption by co-opted CV models leads to information loss. Given this variety of remote sensing modalities, the workshop will focus on strategies for learning while leveraging multiple EO data sources effectively. We invite discussions that address data challenges related to heterogeneity, spatial and temporal resolution, satellite view angle, data fusion and representation formats.
Enable scaling of CV4EO applications: Current EO satellite constellations collect 100+TBs of data a day, and images can be billions of pixels large. While these volumes impose challenges for data management and require customization of model training pipelines, the volume and diversity of remote sensing archival data represent a great potential for a wide variety of applications and the development of Large Vision-Langauge Models.
Evaluation & benchmarking of CV4EO (Large Vision-Language) models: Establishing best practices, benchmarking datasets and standardized evaluation protocols remains an important challenge that, while not unique to CV4EO, is particularly critical for applications with real-world impact. For example, generalization across geographies, seasonalities, imaging conditions, as well as uncertainty quantification are crucial attributes toward robust and trustworthy models suitable for supporting crisis management (e.g. disaster response) and urban/environmental planning workflows. Recent developments on general purpose Foundation Models for EO have been demonstrating great potential to enable improved and novel CV4EO applications, but also underscoring the need for a joint community effort on establishing benchmarks and evaluation practices that go beyond traditional task-specific benchmarks.
Benefit the Computer Vision Community: Tackling the spatial-temporal awareness, data volumes and multimodal reasoning challenges in CV for EO also has broader implications for the CV community. Similar challenges are faced in domains such as autonomous navigation and biomedical image analysis. The workshop will explore how advancements in CV for EO can contribute to addressing these shared challenges.
Foster Talent Formation and Recruitment: The workshop aims to attract and engage talents in CV for EO, as institutions conducting geospatial data analysis face fierce competition for CV and AI expertise. By introducing attendees to applied research opportunities and practical applications, the workshop will contribute to talent formation and recruitment efforts.
Keynote: Flexible, multi-modal foundation models for satellite Earth observations
Bio: Hannah Kerner is an Assistant Professor in the School of Computing and Augmented Intelligence at Arizona State University. Her research focuses on advancing the foundations and applications of machine learning to foster a more sustainable, responsible, and fair future for all. As the AI Lead for NASA’s agriculture programs, NASA Harvest and NASA Acres, she is deploying research methods in real applications across the globe; her projects have directly resulted in optimized agricultural planning, disaster response, and financial relief in various regions around the world. The impact of Kerner’s research was recognized in Forbes 30 Under 30 and the International Research Centre On Artificial Intelligence’s Top 10 projects solving problems related to the UN’s Sustainable Development Goals with AI.
For more: https://hannah-rae.github.io/
Keynote: The transformative role of geospatial technologies in agriculture: how advanced sensing and AI-driven data analysis are reshaping crop improvement, management and sustainability efforts.
Bio: Dr. Shakoor is an Assistant Member and Principal Investigator at the Donald Danforth Plant Science Center, as well as the co-founder and CEO of Agrela Ecosystems. Nadia’s career has centered on integrating cutting-edge geospatial and phenotyping technologies to advance agricultural practices. At the Danforth Center, she has been a driving force behind several high-impact projects, including the Bill and Melinda Gates-funded Sorghum Genomics Toolbox, which developed genomic and phenotyping tools for sorghum breeding, and the ARPA-E funded TERRA-REF project, deploying the world’s largest field crop analytics robot in Maricopa, Arizona. As the founder of Agrela Ecosystems, Nadia has developed the PheNode, a core sensor platform that gathers critical agricultural data, and leads the USDA-funded FieldDock project, which integrates autonomous drones and wireless sensor networks for enhanced field monitoring. She is also collaborating with the Salk Institute for Biological Studies to leverage geospatial sensing technologies for optimizing carbon capture and sequestration in sorghum. Additionally, Nadia is working with the National Sorghum Producers on their USDA-funded Climate Smart-Commodities program, advancing climate-smart sorghum genetics and field management practices.
For more: https://www.danforthcenter.org/our-work/principal-investigators/nadia-shakoor/
We welcome submissions of full papers as well as position papers, work-in-progress, and papers discussing open problems and challenges. Topics include, but are not limited to:
We plan a half-day program, including 2 keynote addresses, individual presentations, and panel discussion. The planned format will consist of lightning talks + poster session, with two main manuscript submission types: