P
Peter M. Roth
Researcher at Graz University of Technology
Publications - 134
Citations - 7517
Peter M. Roth is an academic researcher from Graz University of Technology. The author has contributed to research in topics: Object detection & Video tracking. The author has an hindex of 28, co-authored 132 publications receiving 6694 citations. Previous affiliations of Peter M. Roth include Technische Universität München.
Papers
More filters
Proceedings ArticleDOI
Large scale metric learning from equivalence constraints
TL;DR: This paper introduces a simple though effective strategy to learn a distance metric from equivalence constraints, based on a statistical inference perspective, which is orders of magnitudes faster than comparable methods.
Proceedings ArticleDOI
Annotated Facial Landmarks in the Wild: A large-scale, real-world database for facial landmark localization
TL;DR: AFLW provides a large-scale collection of images gathered from Flickr, exhibiting a large variety in face appearance as well as general imaging and environmental conditions, and is well suited to train and test algorithms for multi-view face detection, facial landmark localization and face pose estimation.
Book ChapterDOI
Person re-identification by descriptive and discriminative classification
TL;DR: The proposed approach is demonstrated on two datasets, where it is shown that the combination of a generic descriptive statistical model and a discriminatively learned feature-based model attains considerably better results than the individual models alone.
Book ChapterDOI
Relaxed pairwise learned metric for person re-identification
TL;DR: This paper proposes to learn a metric from pairs of samples from different cameras, so that even less sophisticated features describing color and texture information are sufficient for finally getting state-of-the-art classification results.
Proceedings ArticleDOI
Hough-based tracking of non-rigid objects
TL;DR: Online learning has shown to be successful in tracking of previously unknown objects, however, most approaches are limited to a bounding-box representation with fixed aspect ratio and cannot handle highly non-rigid and articulated objects.