Lightweight 2D Map Construction of Vehicle Environments Using a Semi-Supervised Depth Estimation Approach

Abstract

This paper addresses the problem of constructing a real-time 2D map for driving scenes from a single monocular RGB image. We presented a method based on three neural networks (depth estimation, 3D object detection, and semantic segmentation). We proposed a depth estimation neural network architecture that is fast and accurate in comparison with the state-of-the-art models. We designed our model to work in real time on light devices (such as an NVIDIA Jetson Nano and smartphones). The model is based on an encoder–decoder architecture with complex loss functions, i.e., normal loss, VNL, gradient loss (dx, dy), and mean absolute error. Our results show competitive results in comparison with the state-of-the-art methods, as our method is 30 times faster and smaller.

Keywords

• machine learning
• deep learning
• computer vision
• monocular depth estimation
• real-time depth estimation
• driver assistant systems