Plans for the Sabbatical at ISL

Prof. Dr.-Ing. Uwe Hahne

Introduction

  • General approach: Working in 3D
    • We live in a 3D world, but most media is 2D
    • Goal: bridge this gap in a way that is useful for the real world
    • Focus areas: 3D perception, 3D reconstruction, 3D interaction
  • Planned focus for the sabbatical at ISL: event-based imaging and its applications for autonomous drones
  • Generative AI or Agentic AI appear to be game changers.

About Me

  • Professor for Computer Vision in media applications at Furtwangen University (since Sept. 2021)
  • Personal site: uhahne.github.io
  • Project overview: uhahne.github.io/projects
  • Fascination since studies: processing, representation and perception of 3D data
  • Career spans academia (TU Berlin), industry (SICK AG) and academia again (HFU)

Academic Background

  • 2006: Diploma "Media Systems" (Diplom-Mediensystemwissenschaftler), Bauhaus University Weimar
  • 2006–2012: PhD student, Computer Graphics Group of Prof. Marc Alexa, Technische Universität Berlin
  • May 2012: PhD thesis "Real-time depth imaging"
    • Advisors: Marc Alexa, Andreas Kolb
    • Dealing with 3D cameras in copmuter graphics applications like Augmented Reality and 3D interaction.

Research during PhD: 3D Interaction

  • Building a multi-touch table for 3D interaction
  • Explored applications in Computer Graphics and Human-Computer Interaction
  • Hahne, U., Schild, J., Elstner, S. & Alexa, M. — Multi-touch focus+context sketch-based interaction, Eurographics SBIM 2009

Research during PhD: Time-of-Flight & 3D Interaction

  • PhD research: real-time depth image processing with time-of-flight cameras, handling noisy/incomplete data using Computer Vision techniques
  • Explored applications in Augmented Reality: Hahne, U. & Alexa, M. — Depth Imaging by Combining Time-of-Flight and On-Demand Stereo, Dynamic 3D Imaging, Springer 2009

Industry Career — SICK AG

  • July 2012: Joined SICK AG in Waldkirch, 3D Compact Systems team, as software development engineer
  • 2017: Became head of the Software Design group
    • Supported multiple research projects at SICK AG, including supervision of students and interns
    • Core work: software development for 3D snapshot cameras
  • Major challenge: synchronizing hardware and software development for a new technology and product line, while also exploring new applications and research directions

Professor at Furtwangen University

  • Since September 2021: Professor for Computer Vision in Media Applications at Furtwangen University
  • Focus: the future of digital 3D media
  • Brings current research topics directly into teaching; involves students in active research projects
  • Builds interactive prototypes with students (e.g. Smart Mirror, multi-touch table)

Teaching at Hochschule Furtwangen — Overview

  • Good teaching is a key ambition — see the teaching portfolio
  • Qualified certificate for university didactics (earned at TU Berlin)
  • Courses: Computer Vision and Deep Learning (CVDL), 3D Computer Vision (3DCV)
  • Computer Graphics (WebGL), Digital Audio/Video Technology (DAVT)
  • New seminar: "KI in Medienanwendungen" (Generative AI in media applications)
  • Basic courses: web programming, physical computing, physics simulation, Math and Coding

IMeRT Courage Hub

  • Founding member of the "Immersive Media in Research and Teaching (IMeRT) Courage Hub"
  • Connects researchers and teachers across HFU locations (Furtwangen, Schwenningen, Freiburg, Tuttlingen)
  • Topics: Mixed Reality, Virtual Reality, Augmented Reality, 3D audio environments
  • Central point of contact for industry partners seeking XR collaboration

Research topics at HFU - NeRF

  • Master thesis basis: Sabine Schleise — NeRFs in the context of the Industrial Metaverse, with SICK AG
  • Implementation and evaluation of a NeRF-based 3D reconstruction pipeline for industrial applications
  • Schleise, S., Hahne, U. & Lindinger, J. — An automated Pipeline to bring NeRFs to the Industrial Metaverse, WSCG 2024

Research topics at HFU - Free3D

Free-viewpoint 3D video creation

  • 3D scene capture from only three Azure Kinect cameras
  • Goal: Neural Radiance Fields (NeRF) or 3D Gaussian Splatting to build a "4D data structure" for novel-view video
  • Challenge: high visual quality with minimal equipment, e.g. for immersive online learning
  • Presented at the European Machine Vision Forum 2024 (Mulhouse)

Research topics at HFU - Smart Mirror

AI-Enhanced Human-Size Interactive Display

  • Camera-based virtual mirror reflecting a user's full-body motion onto a life-size digital avatar in real time
  • Markerless Human Pose Estimation (no wearable sensors)
  • Paper at URAI2026: Fuchs, E. et al. — Human size interactive mirror for functional diagnosis and visual feedback in physiotherapy
  • Outlook: physiotherapy application and event-based imaging

Reflection on a Winding Career

  • Early PhD ambition: publish at each major conference in Computer Graphics, Computer Vision and HCI (SIGGRAPH, CVPR, CHI) — got rejected at all three, but learned a lot about each community
  • Moved to industry (SICK AG) after becoming a father — wanted product impact and stability
  • Missed academic freedom to explore ideas for their own sake
  • Back at HFU: reflecting on teaching undergraduates vs. R&D experts, and on the fast-changing ML/CV research landscape and most recently agentic AI

Research Philosophy

Although we live in a three-dimensional environment, most media are presented on two-dimensional output devices.

Driving questions:

  • How can we capture our 3D world?
  • How can we make it accessible for various applications?
  • How do we handle the loss of information when projecting 3D data onto 2D surfaces?
  • How can we restore 3D information from 2D images?

Current Research Topics

  • Utilizing event-based and all kind of cameras for autonomous drones as well as for human-computer interaction
  • New wave: 3D geometric foundation models (e.g. DepthAnything3, VGGT-Omega) reconstruct 3D scenes directly from 2D images
  • Goal: make these 3D spaces accessible and intuitive to use; open questions around UX, data protection and privacy

Plan for the Sabbatical at ISL

  • No teaching duties until March 2027
  • Sharing: 20 years of experience in Computer Vision and Graphics and do consulting for ISL research projects
  • Learning: explore new research directions and applications in the context of autonomous drones as well as new AI based workflows
  • Topic (but not limited to): event-based imaging and its applications in the context of autonomous drones

Contact & Links