Guided Depth Completion with Instance Segmentation Fusion in Autonomous Driving Applications.

Pixel-level depth information is crucial to many applications, such as autonomous driving, robotics navigation, 3D scene reconstruction, and augmented reality. However, depth information, which is usually acquired by sensors such as LiDAR, is sparse. Depth completion is a process that predicts missing pixels' depth information from a set of sparse depth measurements. Most of the ongoing research applies deep neural networks on the entire sparse depth map and camera scene without utilizing any information about the available objects, which results in more complex and resource-demanding networks. In this work, we propose to use image instance segmentation to detect objects of interest with pixel-level locations, along with sparse depth data, to support depth completion. The framework utilizes a two-branch encoder-decoder deep neural network. It fuses information about scene available objects, such as objects' type and pixel-level location, LiDAR, and RGB camera, to predict dense accurate depth maps. Experimental results on the KITTI dataset showed faster training and improved prediction accuracy. The proposed method reaches a convergence state faster and surpasses the baseline model in all evaluation metrics.

Managing active cultural differences in U.S. construction workplaces: Perspectives from non-Hispanic workers.

INTRODUCTION: Current census reports indicate a growing shift toward workforce diversity in the U.S. construction industry, which is largely the result of increasing participation from the Hispanic community. The data also suggest that the Hispanic workforce suffers a higher rate of fatal injuries compared to their non-Hispanic counterparts. Therefore, there is a dire need to develop and utilize new management tools and strategies to accommodate the differences in language and culture of this incoming labor force. METHOD: The absence of these tools and strategies poses several challenges including cost overrun, schedule delay, and more importantly, higher workplace injury rates. This study aims to provide a better understanding of the contribution of cultural diversity as a factor that may influence the overall site safety. RESULTS: As a result, this study provides further evidence that indicate that the current findings regarding the influence of active cultural differences are reliable, valid, and needs attention. Furthermore, the study provides sub-analysis results of cultural values among Hispanic workers, which suggest that workers from Mexico are less likely to speak up on safety issues when compared to other Hispanic workers. Therefore, this study has both practical and theoretical implications for managing workforce diversity and related safety performance in the U.S. construction industry. The results of the study can be used by employers and managers to adopt responsive strategies and tools to reduce the likelihood of fatal and nonfatal injuries among Hispanic workers.