Explore Insider perspectives

Discover a range of answers from our Insiders, shaped by their roles, journeys and experiences at Nokia.

Hello Andras, I’m a senior tech executive currently specializing in AI Strategy (MIT) and developing a 'Digital Windows' project for transport infrastructure. I’m applying for the Visual AI intern position and would love to know how Nokia balances pure p

Jeremie S. asked a question to András L.

I am currently applying for the Visual AI Systems Research position. With a background in CTO leadership and a current project focused on digital infrastructure for the RATP (Paris Metro), I am particularly interested in how Nokia is scaling neural video compression for urban mobility in France. I would love to briefly ask: How does the team balance pure academic research with the hardware constraints of industrial deployment?Looking forward to your insights.Best regards, Jérémie

View favourites
  • 1 reply
  • 1 views
  • Author: Jeremie S.
  • Category: Eligibility advice, Qualifications
  • Date asked:
  • Last update:
  • András L.
    András L. Sw Dev. Specialist

    Hi,

    I am working on a different area, so perhaps other colleagues could provide more detailed insights. However, Nokia generally addresses the balance between academic research and hardware constraints for neural video compression in urban mobility by focusing on efficient model architectures and hardware-aware algorithm design from the outset. This involves co-optimizing compression algorithms with target edge processing units and network capabilities, often leveraging techniques like quantization, pruning, and neural architecture search to reduce computational complexity. For urban mobility in France, this translates to developing solutions that can operate reliably on existing or planned infrastructure, ensuring real-time performance and data efficiency. We prioritize research into low-latency, high-fidelity compression methods that are robust to varying network conditions and camera types typical in public transport environments. This strategic approach allows for the practical deployment of advanced AI while maintaining operational feasibility and cost-effectiveness. Our goal is to bridge the gap between theoretical advancements and real-world application through iterative development and rigorous testing.

    Best regards,

    LA