DVPS
A European research initiative coordinated by Translated to develop next-generation multimodal AI systems that learn through real-world interaction.
DVPS operates at the intersection of geo intelligence and artificial intelligence, focusing on the fusion of diverse spatial, sensor, and textual data into multimodal AI systems. The company serves institutional clients such as the European Commission, the European Space Agency, and the United Nations, providing solutions for smarter disaster response, urban planning, and environmental forecasting. DVPS's business model revolves around large-scale projects, often valued in the millions of euros, where they leverage their expertise in deep learning and remote sensing. The company is funded by the European Horizon Europe Framework, which supports their research and development efforts. DVPS's multimodal approach allows for the interpretation of complex real-world scenarios, overcoming the limitations of single-modal AI systems. Their models are designed to be label-efficient, capable of learning from limited labeled data through transfer learning and few-shot adaptation. This positions DVPS as a key player in the development of foundation models for mining information from multimodal Earth observation data. The company also emphasizes the ethical implications of AI, ensuring robust and responsible AI development. Keywords: geo intelligence, multimodal AI, disaster response, urban planning, environmental forecasting, deep learning, remote sensing, institutional clients, foundation models, ethical AI.