ICCV (International Conference on Computer Vision) is the premier international computer vision event comprising the main conference and several co-located workshops and tutorials.
ICCV2017 was held at the Venice Convention Center in Venice, Italy and gathered computer vision experts from around the world.
Tomoyuki Mukasa, a researcher of Rakuten Institute of Technology published our research at the Geometry Meets Deep Learning ICCV 2017 Workshop.
Our paper “3D Scene Mesh From CNN Depth Predictions And Sparse Monocular SLAM” proposed a method integrating geometrical measurements of visual SLAM and CNN depth prediction by mesh deformation and obtain a more accurate dense 3D reconstruction with scale. We further propose a system architecture suitable for mobile devices, where feature tracking and CNN-based depth prediction modules are separated.