You are here

Oswald Lanz

Permanent researcher
  • Phone: 0461314511
  • FBK Povo
Short bio

I am Research Scientist at Fondazione Bruno Kessler - FBK and former head of the computer vision research Unit TeV of FBK.

My research belongs to the area of computer vision and machine learning and is centered on methods and systems for video understanding and 3d scene analysis. I have worked extensively on object tracking, audio-visual trackinghead pose estimation and group detection for behavior analysis. My current focus is on representation learning for video.


2020.07: News featured on FBK Magazine.

2020.06: Receiving the AWS Machine Learning Research Award.

2020.06: We ranked 3rd place in the CVPR'20 EPIC Kitchens Action Recognition challenge.

2020.02: Gate-Shift Networks (GSM) accepted at CVPR'20 main program.

2020.02: We are organizing a Workshop on Image and Video Question Answering.

2019.12: GSM is state of the art in action recognition.

2019.12: Journal Paper accepted at IEEE Transactions on Image Processing.

2019.11: I gave an invited talk at ICCV'19 Egocentric Perception, Interaction and Computing workshop in Seoul, Korea.

2019.11: Let's welcome Alex, new PhD student on the topic of video question answering.

2019.07: Paper accepted at ICCV'19 main program. 

2019.07: I gave a seminar Learning to Recognize Actions in Video at Siena Artificial Intelligence Lab.

2019.06: Swathikiran defended his PhD thesis and graduated with cum laude mark.

2019.06: We ranked 3rd place in the CVPR'19 EPIC Kitchens Action Recognition challenge. Tech Report available here.

2019.06: I gave a seminar on our work on action recognition at FBK Spring of AI seminar series.

2019.04: Paper accepted at ICIP'19 main program.

2019.02: Long Short-Term Attention (LSTA) accepted at CVPR'19 main program.

2019.02: Paper accepted at ICASSP'19 main program.

2019.01: Journal Paper accepted at ACM MultiMedia.

2019.01: Journal Paper accepted at Intelligenza Artificiale special issue on selected papers of AI*IA 2018 conference.

  1. O. Lanz,
    Occlusion Robust Tracking of Multiple Objects,
    Computational Imaging and Vision,
    , pp. 715-
    , (International Conference on Computer Vision and Graphics - ICCVG,
    Warsaw (Poland),
    from 09/22/2004 to 09/24/2004)
  2. O. Lanz,
    Modeling Interactions in Multiple Object Bayesian Tracking,
    This paper proposes a framework for modeling interactions in muliple object 3D Bayesian tracking. It exploits both the computational cheapness of independent single object filters and the modeling power of the joint filter. Sampling from complex joint propagation densities is avoided by introducing interaction a posteriori, after blind single object propagation has been performed. To deal with occlusions for each object a support layer is computed. It contains probabilistic information about how likely a pixel is occluded by another object. Utilized to give less weight to likely occluded pixels, it provides the basis of a robust likelihood model. The implementation of the proposed ideas in a Sequential Monte Carlo framework are discussed. Experiments on a synthetic data show the robustness of the proposed ideas,
  3. Alessandro Santuari; Oswald Lanz; Roberto Brunelli,
    Synthetic Movies for Computer Vision Applications,
    Proceedings of the Third IASTED International Conference on Visualization, Imaging, and Image Processing,
    , pp. 139-
  4. Filippo Bertamini; Roberto Brunelli; Oswald Lanz; Alessandro Roat; F. Santuari; Q. Xu Tobia,
    Olympus: an Ambient Intelligence Architecture on the Verge of Reality,
    Proceedings of the 12th International Conference on Image Analysis and Processing [ICIAP 2003],
    , pp. 139-
  5. Filippo Bertamini; Roberto Brunelli; Oswald Lanz; Alessandro Roat; Alessandro Santuari; Francesco Tobia; Qing Xu,
    Olympus: an ambient intelligence architecture on the verge of reality,
  6. Alessandro Santuari; Oswald Lanz; Roberto Brunelli,
    Synthetic Movies for Computer Vision Applications,
    This paper presents a real time graphical simulator based on a client-server architecture. The rendering engine, supported by a specialized client application for the automatic generation of goal oriented motion of synthetic characters, is used to produce realistic image sequences for extensive performance assessment of computer vision algorithms,