Milestone unveils Project Hafnia for AI video data access

Milestone Systems has announced Project Hafnia, a platform designed to facilitate access to video data and the training of AI models using NVIDIA Cosmos Curator and AI model fine-tuning microservices.

The project aims to address the challenge often faced by AI developers: obtaining adequate, reliable video data necessary for training visual AI models.

By leveraging the NVIDIA platform, Milestone Systems' Project Hafnia intends to service data generators keen to share their data and developers seeking traceable, regulatory-compliant annotated video data.

Milestone Systems is positioning itself to distribute the latest advances in Vision Language Models (VLM) and supporting data curation capabilities more widely among developers.

A key feature within their service offerings is a VLM fine-tuned via NVIDIA microservices on a substantial amount of compliant transportation data, curated with the help of NVIDIA Cosmos Curator.

The quality and effectiveness of the VLMs are enhanced for operation on NVIDIA GPUs and utilise NVIDIA's video search and summarisation (VSS) AI blueprint.

Deepu Talla, NVIDIA Vice President and General Manager of Embedded and Edge Computing, commented on this collaboration, "By leveraging the NVIDIA platform, Milestone Systems is helping accelerate this next wave of powerful visual services. The next phase in development and adoption of visually perceptive Agentic AI services will be unlocked by recipes like NVIDIA VSS blueprint combined with widely available and accessible fine-tuned VLM models."

Thomas Jensen, CEO of Milestone Systems, highlighted the importance of Project Hafnia, stating that the main obstacle in AI development is the availability of quality data for training purposes. "Artificial intelligence is our generation's biggest game-changer."

"A major challenge for the ongoing development is having access to enough high-quality data for training AI models."

"The Project Hafnia platform will collect and curate data with the aspiration to be the world's smartest, fastest and responsible platform for video data and training of AI models," he said.

Project Hafnia aspires to resolve major obstacles by streamlining data access. NVIDIA Cosmos Curator's data curation will expedite AI and analytics development, enhancing processing speeds to be up to 30 times quicker than current norms.

Milestone Systems has outlined two key services: Training as a Service for AI model development using high-quality video data and a Visual Language Model as a Service for smart city transportation and Intelligent Traffic Service use cases. These services aim to provide industry-leading performance.

The project offers new prospects for the development of high-quality data and annotation techniques that could improve future analytic software, potentially leading to large-scale automation in areas such as traffic management, manufacturing, airports, law enforcement, and business sectors.

Milestone Systems has already initiated steps to establish a traffic and transportation-specific Visual Language Model (VLM) through NVIDIA's platform.

This model addresses applications like general traffic evaluations, driving condition assessments, alert validation, and incident reporting.

Project Hafnia is currently in its pilot phase, allowing interested developers an opportunity to register for early access. Initially, the platform will focus on traffic data, with plans to expand its range once it is fully operational. Milestone Systems both owns and manages the platform.