You are here

Oswald Lanz

Permanent researcher
  • Phone: 0461314511
  • FBK Povo
Short bio

I received the MSc in Mathematics ('00) and a Ph.D in Computer Science ('05) from the University of Trento, and the Habilitation to Full Professor (in Italy, ASN16 09/H1).

I am a Researcher at the Fondazione Bruno Kessler, Trento, Italy since 2005, tenured since 2008. My research has been mainly in the field of people tracking, audio-visual tracking, pose estimation and group detection with application to monitoring and behavior analysis. From 2012 to end of 2015 I have been Head of Computer Vision Research Unit Technologies of Vision at the Fondazione Bruno Kessler. In 2016 I have been launching an innovation project in a home automation SME. 

My main research interests are currently in deep learning with structured output prediction applied to video classification and tracking. I hold a number of international patents on video tracking.

Publications
  1. O. Lanz,
    Method and apparatus for tracking a number of objects or object parts in image sequences.
    2007,
    A method for tracking a number of objects or object parts in image sequences utilizes a Bayesian-like approach to object tracking, computing, at each time a new image is available, a probability distribution over all possible target configurations for that time. The Bayesian-like approach to object tracking computes a probability distribution for the previous image, at time (t−1), is propagated to the new image at time (t) according to a probabilistic model of target dynamics, obtaining a predicted distribution at time (t). The Bayesian-like approach to object tracking also aligns the predicted distribution at time (t) with the evidence contained in the new image at time (t) according to a probabilistic model of visual likelihood.
  2. O. Lanz,
    Approximate Bayesian Multibody Tracking,
    in «IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE»,
    vol. 28,
    n. 9,
    2006
    , pp. 1436 -
    1449
  3. Roberto Brunelli; Alessio Brutti; Paul Chippendale; Oswald Lanz; Maurizio Omologo; Piergiorgio Svaizer; Francesco Tobia,
    A Generative Approach to Audio-Visual Person Tracking,
    Multimodal Technologies for Perception,
    Springer,
    vol.4122/2007,
    2006
    , (First International Evaluation Workshop on Classification of Events, Activities and Relationships - CLEAR'06,
    Southampton, UK,
    from 04/10/2006 to 04/12/2006)
  4. O. Lanz; F. Tobia; R. Brunelli,
    Multi-view Appearance Model for Visual People Tracking,
    Atti della Seconda Conferenza Italiana sui Sistemi Intelligenti - CISI,
    2006
    , (Seconda Conferenza Italiana sui Sistemi Intelligenti - CISI 2006,
    Ancona, Italy,
    09/27/2006 - 09/29/2006)
  5. O. Lanz; R. Brunelli,
    Dynamic Head Location and Pose from Video,
    IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems 2006,
    n. 4042078,
    2006
    , pp. 47-
    52
    , (IEEE Conference on Multisensor Fusion and Integration - MFI,
    Heidelberg (Germany),
    da 09/03/2006 a 09/06/2006)
  6. O. Lanz,
    Method and apparatus for tracking a number of objects or object parts in image sequence.
    2006,
    The present invention provides for a method for tracking a number of objects or object parts in image sequences, comprising following a Bayesian-like approach to object tracking, computing, at each time a new image is available, a probability distribution over all possible target configurations for that time, said Bayesian-like approach to object tracking comprising the following steps:
    - Prediction step: a probability distribution is computed for the previous image, at time (t-1), is propagated to the new image at time (t) according to a probabilistic model of target dynamics, obtaining a predicted distribution at time (t) ;
    - Update step: the predicted distribution at time (t) is then aligned with the evidence contained in the new image at time (t) according to a probabilistic model of visual likelihood.
  7. O. Lanz; R. Manduchi,
    Hybrid Joint-Separable Multibody Tracking,
    vol.1,
    2005
    , pp. 413-
    420
    , (IEEE Conference on Computer Vision and Pattern Recognition - CVPR,
    San Diego (CA),
    from 06/20/2005 to 06/26/2005)
  8. O. Lanz,
    Probabilistic Multi-Person Tracking for Ambient Intelligence,
    Monitoring human activities with visual sensors is still a challenge, especially when multiple targets are involved. Occlusions, if not properly handled, are a major source of failure. Indoor environments with complex topology require the use of sensor networks whose effective management is by itself a difficult problem. Situated in the context of Ambient Intelligence, this thesis is concerned with both algorithmic and architectural aspects of distributed monitoring systems. The algorithmic approach pursued is that of non–parametric Bayesian filtering, a probabilistic state estimation framework whose multi target formulation allows physically–based modeling of the occlusion process within an appearance based, background independent observation model. While unaffordable in its plain, joint formulation, a novel, efficient, probabilistically sound solution is proposed and its robustness experimentally verified. Its MonteCarlo implementation provides a suitable framework for tackling dynamic resource allocation within a distributed modular architecture. Representation size of probabilistic estimates is adapted to conveyed uncertainty, while active environment sampling aims at minimizing uncertainty. Independent sensor agencies are synergically reconfigured by a supervisor process with global environment knowledge. The proposed system is then adaptive to its own performance, environment and sensing and computing infrastructure, providing a scalable solution to the problem of visual monitoring of crowded, topologically complex environments,
    2005
  9. O. Lanz,
    A Joint-Separable Filter for Multitarget Tracking,
    Tracking multiple targets with visual sensors is a challenging problem, especially when efficiency is an issue. Occlusions, if not properly handled, are a major source of failure. Solutions supporting principled occlusion reasoning have been proposed but are yet unpractical for online applications. This article presents a new solution which effectively manages the trade-off between reliable modeling and computational efficiency. The Hybrid Joint-Separable (HJS) filter is derived from a joint Bayesian formulation of the problem, and shown to be efficient while optimal in terms of efficient belief representation. Computational efficiency is achieved by employing a forward MRF approximation to joint dynamics and an incremental algorithm for posterior update with an appearance likelihood that implements a physically-based model of the occlusion process. Its real time MonteCarlo implementation achieves accurate tracking during partial occlusions, while in case of complete occlusion tracking hypotheses are bound to estimated occlusion volumes. Experiments show that the proposed algorithm is very robust and able to resolve even long term complete occlusions between targets with identical appearance,
    2005
  10. O. Lanz,
    Dynamic Resource Allocation for Probabilistic Tracking in Complex Scenes,
    Monitoring human activities in large environments is a challenging problem. Such scenarios impose the use of sensor networks and decentralized processing to achieve scalability in terms of spatial environment and task complexity. This article addresses both algorithmic and architectural aspects of distributed tracking systems. The algorithmic approach pursued is that of sequential Bayesian filtering, whose standard formulation is rewritten in a form suitable for distribution within a three-layered modular architecture. Its non-parametric MonteCarlo implementation provides a suitable framework for tackling dynamic resource allocation: representation size of probabilistic estimates is adapted to conveyed uncertainty, while active environment sampling aims at minimizing uncertainty. Independent tracking agencies with competence on a subset of targets are instantiated and continuously reconfigured by a supervisor process with global environment knowledge. The proposed system is then adaptive to its sensing and computing infrastructure, own performance and environment, providing a scalable solution to the problem of visual monitoring of populated, topologically complex environments,
    2005

Pages