J
Jan Cech
Researcher at Czech Technical University in Prague
Publications - 39
Citations - 968
Jan Cech is an academic researcher from Czech Technical University in Prague. The author has contributed to research in topics: Convolutional neural network & Stereo camera. The author has an hindex of 15, co-authored 39 publications receiving 846 citations. Previous affiliations of Jan Cech include Chinese Academy of Sciences & French Institute for Research in Computer Science and Automation.
Papers
More filters
Proceedings Article
Visual Heart Rate Estimation with Convolutional Neural Network.
TL;DR: A novel two-step convolutional neural network is proposed to estimate a heart rate from a sequence of facial images to test the robustness of heart rate estimation methods to illumination changes and subject’s motion.
Proceedings ArticleDOI
Scene flow estimation by growing correspondence seeds
TL;DR: A simple seed growing algorithm for estimating scene flow in a stereo setup that is accurate for complex scenes with large motions and produces temporally-coherent stereo disparity and optical flow results is presented.
Proceedings ArticleDOI
Efficient Sampling of Disparity Space for Fast And Accurate Matching
Jan Cech,Radim Sara +1 more
TL;DR: A simple stereo matching algorithm is proposed that visits only a small fraction of disparity space in order to find a semi-dense disparity map by growing from a small set of correspondence seeds, which is very unlike the existing growing algorithms which are fast but erroneous.
Journal ArticleDOI
Efficient sequential correspondence selection by cosegmentation
TL;DR: It is shown experimentally that the proposed sequential correspondence verification (SCV) algorithm significantly outperforms the standard correspondence selection method based on SIFT distance ratios on challenging matching problems.
Proceedings ArticleDOI
Topologically-robust 3D shape matching based on diffusion geometry and seed growing
TL;DR: This paper proposes to use a shape descriptor based on properties of the heat-kernel which provides an intrinsic scale-space representation and shows that it can deal with substantial topological differences between the two shapes.