M
Michael Bleyer
Researcher at Microsoft
Publications - 71
Citations - 4298
Michael Bleyer is an academic researcher from Microsoft. The author has contributed to research in topics: Image segmentation & Depth map. The author has an hindex of 22, co-authored 71 publications receiving 3914 citations. Previous affiliations of Michael Bleyer include Vienna University of Technology.
Papers
More filters
Proceedings ArticleDOI
Fast cost-volume filtering for visual correspondence and beyond
TL;DR: This paper proposes a generic and simple framework comprising three steps: constructing a cost volume, fast cost volume filtering and winner-take-all label selection, and achieves state-of-the-art results that achieve disparity maps in real-time, and optical flow fields with very fine structures as well as large displacements.
Proceedings ArticleDOI
PatchMatch Stereo - Stereo Matching with Slanted Support Windows.
TL;DR: The method reconstructs highly slanted surfaces and achieves impressive disparity details with sub-pixel precision and allows for explicit treatment of occlusions and can handle large untextured regions.
Journal ArticleDOI
Fast Cost-Volume Filtering for Visual Correspondence and Beyond
TL;DR: This work proposes a generic and simple framework comprising three steps: constructing a cost volume, fast cost volume filtering, and 3) Winner-Takes-All label selection that achieves 1) disparity maps in real time whose quality exceeds those of all other fast (local) approaches on the Middlebury stereo benchmark, and 2) optical flow fields which contain very fine structures as well as large displacements.
Proceedings ArticleDOI
Local stereo matching using geodesic support weights
TL;DR: The proposed algorithm is the top performer among local stereo methods at the current state-of-the-art in local stereo matching by using the geodesic distance transform.
Proceedings ArticleDOI
Object stereo — Joint stereo matching and object segmentation
TL;DR: This paper presents a method for joint stereo matching and object segmentation that is able to recover the depth of regions that are fully occluded in one input view, which to the knowledge is new for stereo matching.