Deep MANTA: A Coarse-to-Fine Many-Task Network for Joint 2D and 3D Vehicle Analysis from Monocular Image

• Published in: 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) pp. 1827 - 1836
• Main Authors: Chabot, Florian, Chaouch, Mohamed, Rabarisoa, Jaonary, Teuliere, Celine, Chateau, Thierry
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.07.2017
• Subjects: Object detection; Pose estimation; Proposals; Shape; Solid modeling; Three-dimensional displays; Two dimensional displays

In this paper, we present a novel approach, called Deep MANTA (Deep Many-Tasks), for many-task vehicle analysis from a given image. A robust convolutional network is introduced for simultaneous vehicle detection, part localization, visibility characterization and 3D dimension estimation. Its archite...