D
David Hafner
Researcher at Saarland University
Publications - 17
Citations - 578
David Hafner is an academic researcher from Saarland University. The author has contributed to research in topics: Anisotropic diffusion & Smoothing. The author has an hindex of 8, co-authored 17 publications receiving 405 citations. Previous affiliations of David Hafner include Intel.
Papers
More filters
Journal ArticleDOI
Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-shot Cross-dataset Transfer.
TL;DR: The authors proposed a robust training objective that is invariant to changes in depth range and scale, advocate the use of principled multi-objective learning to combine data from different sources, and highlight the importance of pretraining encoders on auxiliary tasks.
Book ChapterDOI
Why Is the Census Transform Good for Robust Optic Flow Computation
TL;DR: This paper studies the continuous limit of the inherently discrete census transform and embeds it into a variational setting and shows two surprising results: the census-based technique enforces matchings of extrema, and it induces an anisotropy in the data term by acting along level lines.
Journal ArticleDOI
The h-index is no longer an effective correlate of scientific reputation
Vladlen Koltun,David Hafner +1 more
TL;DR: In this article, a large-scale study of scientometric measures, analyzing millions of articles and hundreds of millions of citations across four scientific fields and two data platforms, was conducted, finding that the correlation of the H-index with awards that indicate recognition by the scientific community has substantially declined.
Proceedings ArticleDOI
Simultaneous HDR and Optic Flow Computation
TL;DR: This work presents the first approach that simultaneously computes the aligned HDR composite as well as accurate displacement maps and designs a fully coupled model transparently in a well-founded variational framework that can not only cope with dynamic scenes but even precisely represent the underlying motion.
Proceedings ArticleDOI
The Complete Rank Transform: A Tool for Accurate and Morphologically Invariant Matching of Structures
TL;DR: In this straightforward framework, a novel descriptor of local image structure that carries the maximally possible amount of localimage information under morphological invariance is developed and it is demonstrated that this descriptor is preferable over related features that exhibit the same invariance.