Daniel Kurz
Engineer and Manager in Machine Learning, Computer Vision, and Augmented Reality
Daniel's teams at Apple have shipped ARKit features in iOS, delivered large-scale high-fidelity data collection and processing tools for Apple's LiDAR scanner, and introduced 3D scene understanding and product fitting algorithms to visionOS and Vision Pro.
He currently contributes to the development of a modern, LLM-based Siri. Before Apple, Daniel conducted Spatial AR research at universities in Germany and Japan and built Mobile AR algorithms and products at the pioneering AR startup Metaio.
Experience
Apple
- Ran a team focussed on context engineering and model training for LLM-based Siri responses.
- Drove the delivery of visionOS features including environment light estimation for coherently lit virtual content, room analysis for spatial audio, video pass-through improvements for camera regions and surroundings zoom, and the algorithm for finding the right fit for Vision Pro.
- Directed the team designing and building tools for large-scale in-the-wild data collections for 3D scene understanding algorithms, including ARKitScenes and datasets that enabled launching Apple’s LiDAR Scanner in iPad Pro.
- Led a team developing ARKit algorithms enhancing visual coherence and realism. Shipped ambient light estimation, face tracking-enabled directional light estimation, coherent camera noise, and GAN-based realistic reflections in iOS.
- Spearheaded the camera and compute requirements engineering for algorithms on Vision Pro.
Metaio
- Founded and led the Advanced Technologies Group, researching and innovating in Augmented Reality, Computer Vision, Computer Graphics, Sensor Fusion, and Human Computer Interfaces to strengthen Metaio’s IP portfolio and increase visibility as a technology leader, leading to its acquisition.
- Invented new Computer Vision and Sensor Fusion methods for camera localization, addressing outdoor challenges and low-texture objects. Led iOS app development for customers (e.g. Audi eKurzinfo) and Metaio Toolbox. Designed and implemented custom algorithms for Volkswagen, Red Bull, and more.
- Designed, implemented, and optimized novel and efficient Computer Vision algorithms for real-time visual object detection and tracking on mobile devices. These algorithms became integral components of the MetaioSDK.
Education
Technical University of Munich · Prof. Gudrun Klinker, Ph.D.
2012 — 2014Dissertation: Sensor-Aided Visual Camera Localization and Tracking for Handheld Augmented Reality
Bauhaus-University Weimar · Prof. Dr. Ing. Oliver Bimber
2007 — 2010Osaka University · Prof. Kiyoshi Kiyokawa, Ph.D.
2007 — 2008Bauhaus-University Weimar · Prof. Dr. Ing. Oliver Bimber
2003 — 2007Publications and Patents
23 publications and 42 patents with over 2,400 citations. Full lists below; see also Google Scholar.
Publications
- Reconstructing training data from diverse ML models by ensemble inversion Proc. IEEE/CVF Winter Conference on Applications of Computer Vision, 2022 @arXiv
- ARKitScenes: A diverse real-world dataset for 3d indoor scene understanding using mobile RGB-D data NeurIPS Datasets and Benchmarks, 2021 @arXiv
- HDR environment map estimation for real-time augmented reality Proc. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2021 @arXiv
- Leveraging the user's face for absolute scale estimation in handheld monocular SLAM Proc. IEEE International Symposium on Mixed and Augmented Reality (ISMAR), 2016 Preprint @IEEE
- Real-time illumination estimation from faces for coherent rendering Proc. IEEE International Symposium on Mixed and Augmented Reality (ISMAR), 2014 Preprint @IEEE
- Thermal touch: Thermography-enabled everywhere touch interfaces for mobile augmented reality applications Proc. IEEE International Symposium on Mixed and Augmented Reality (ISMAR), 2014 Preprint @IEEE
- [Poster] Towards mobile augmented reality for the elderly Proc. IEEE International Symposium on Mixed and Augmented Reality (ISMAR), 2014 Preprint @IEEE
- [Poster] A Mobile Augmented Reality System to Assist Auto Mechanics Proc. IEEE International Symposium on Mixed and Augmented Reality (ISMAR), 2014 Preprint @IEEE
- [Poster] Smartwatch-aided handheld augmented reality Proc. IEEE International Symposium on Mixed and Augmented Reality (ISMAR), 2014 Preprint @IEEE
- Absolute spatial context-aware visual feature descriptors for outdoor handheld camera localization Proc. International Conference on Computer Vision Theory and Applications (VISAPP), 2014 Preprint @SciTePress
- An outdoor ground truth evaluation dataset for sensor-aided visual handheld camera localization Proc. IEEE International Symposium on Mixed and Augmented Reality (ISMAR), 2013 Preprint @IEEE
- Representative feature descriptor sets for robust handheld camera localization Proc. IEEE International Symposium on Mixed and Augmented Reality (ISMAR), 2012 Preprint @IEEE
- Handheld augmented reality involving gravity measurements Computers & Graphics, 2012 @Elsevier
- Benchmarking inertial sensor-aided localization and tracking methods Proc. TrakMark Workshop (ISMAR), 2011 Preprint
- Gravity-aware handheld augmented reality Proc. IEEE International Symposium on Mixed and Augmented Reality (ISMAR), 2011 Preprint @IEEE
- Inertial sensor-aligned visual feature descriptors Proc. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2011 Preprint @IEEE
- Color invariant chroma keying and color spill neutralization for dynamic scenes and cameras The Visual Computer, 2010 Preprint @Springer
- Fast and robust CAMShift tracking Proc. IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2010 @IEEE
- Closed-loop feedback illumination for optical inverse tone-mapping in light microscopy IEEE Transactions on Visualization and Computer Graphics, 2010 @IEEE
- Projected light microscopy SIGGRAPH 2009: Talks, 2009 @ACM
- Improving Interaction Performance for Ray Tracing Eurographics (Short Papers), 2008 Preprint @Eurographics
- Mutual occlusions on table-top displays in mixed reality applications Proc. ACM Symposium on Virtual Reality Software and Technology (VRST), 2008 Preprint @ACM
- Laser pointer tracking in projector-augmented architectural environments Proc. IEEE and ACM International Symposium on Mixed and Augmented Reality (ISMAR), 2007 Preprint @IEEE
Patents
- Dynamic frame selection for scene understanding US Patent 12,536,737 · 2026
- Specular surface mapping US Patent 12,373,969 · 2025
- Immersive virtual display US Patent 12,347,047 · 2025
- Image-based detection of surfaces that provide specular reflections and reflection modification US Patent 11,900,569 · 2024
- Method and system for presenting a digital information related to a real object US Patent 12,039,644 · 2024
- Transparent insert identification US Patent 11,948,043 · 2024
- Method and device for eye tracking using event camera data US Patent 12,105,280 · 2024
- Method and device for detecting a touch between a first object and a second object US Patent 11,797,132 · 2023
- Global reference direction detection for orienting three-dimensional models US Patent 11,935,263 · 2024
- Method and system for generating at least one image of a real environment US Patent 11,877,086 · 2024
- Light probe placement for displaying objects in 3D environments on electronic devices US Patent 12,002,165 · 2024
- Detecting specular surfaces US Patent 12,073,533 · 2024
- Multiple device model augmentation US Patent Application 18/441,274 · 2024
- Method and system for determining at least one property related to at least part of a real environment US Patent 11,727,661 · 2023
- Method of and system for projecting digital information on a real object in a real environment US Patent 11,652,965 · 2023
- Method and system for determining spatial coordinates of a 3D reconstruction of at least part of a real object at absolute spatial scale US Patent 11,741,624 · 2023
- Environment synthesis for lighting an object US Patent 11,694,392 · 2023
- Partial image completion US Patent 11,636,578 · 2023
- Devices and methods for motion planning of computer characters US Patent Application 18/174,560 · 2023
- Training a network to inhibit performance of a secondary task US Patent 11,734,570 · 2023
- Posture detection and correction US Patent 11,544,865 · 2023
- Classification for image creation US Patent 11,423,308 · 2022
- Immersive display of motion-synchronized virtual content US Patent 11,321,923 · 2022
- Rendering objects to match camera noise US Patent 11,308,652 · 2022
- Deformable object tracking US Patent 11,379,996 · 2022
- Generating a computer graphic for a video frame US Patent 11,282,171 · 2022
- Federated learning using local ground truth estimation US Patent 11,507,836 · 2022
- Postponing the state change of an information affecting the graphical user interface until during the condition of inattentiveness US Patent 11,334,156 · 2022
- Modeling a geographical space for a computer-generated reality experience US Patent 11,488,352 · 2022
- Rendering virtual objects with realistic surface properties that match the environment US Patent 10,930,049 · 2021
- Rendering virtual content with coherent visual properties US Patent 11,182,978 · 2021
- Method of detecting and describing features from an intensity image US Patent 10,607,350 · 2020
- Method of providing image feature descriptors US Patent 10,528,847 · 2020
- Augmented virtual display US Patent 10,825,255 · 2020
- Depth-based touch detection US Patent 10,572,072 · 2020
- Method of providing a descriptor for at least one feature of an image and method of matching features US Patent 10,650,546 · 2020
- Method and system for determining a model of at least part of a real object US Patent 10,297,083 · 2019
- Augmented safety restraint US Patent 10,189,434 · 2019
- Method and system for providing information associated with a view of a real environment superimposed with a virtual object US Patent Application 15/100,355 · 2016
- Method of matching image features with reference features US Patent 9,400,941 · 2016
- Method for registering at least one part of a first and second image using a collineation warping function US Patent 9,213,908 · 2015