Multi-Stage CNN-Based Monocular 3D Vehicle Localization and Orientation Estimation
This work addresses the problem of 3D object detection for autonomous driving systems, offering an incremental approach to improve localization and orientation.
This paper proposes a multi-stage CNN model for 3D vehicle localization and orientation estimation from monocular images. It combines a bird's-eye view elevation map with deep feature representations to estimate object depth and 3D bounding boxes.
This paper aims to design a 3D object detection model from 2D images taken by monocular cameras by combining the estimated bird's-eye view elevation map and the deep representation of object features. The proposed model has a pre-trained ResNet-50 network as its backend network and three more branches. The model first builds a bird's-eye view elevation map to estimate the depth of the object in the scene and by using that estimates the object's 3D bounding boxes. We have trained and evaluate it on two major datasets: a syntactic dataset and the KIITI dataset.