Millimeter-wave Object Recognition Dataset (MORD)

In this study, the researchers employ a commercially available millimeter-wave (MMW) radar to collect data and investigate the effectiveness of deep learning algorithms in recognizing various objects. The study examines the impact of different environmental conditions, including height, distance, and lighting, on object recognition performance for both static and dynamic radar modes. To gather data using the static radar mode, the researchers select five different objects made of four materials - a dime, quarter, lead pencil, plastic sheet, and wood - and test them under three different lighting conditions and at two different heights. In the moving radar mode, the researchers use little bit five different objects made of four materials - a UGV, water bottle, plastic sheet, paper, and clothes - to collect data under three lighting conditions and at two different distances. The primary objective of the study is to identify potential challenges in object recognition using MMW radar and explore approaches to overcome them. By analyzing the collected data, the researchers aim to provide insights into how environmental factors impact object recognition performance and devise strategies to improve the accuracy of deep learning algorithms in recognizing objects. Overall, the study seeks to advance the understanding of MMW radar-based object recognition and contribute to the development of more robust and reliable deep learning algorithms for this purpose.