RF-SLAM: UHF-RFID Based Simultaneous Tags Mapping and Robot Localization Algorithm for Smart Warehouse Position Service

Wu, Chong; Gong, Zeyu; Tao, Bo; Tan, Kong Wai; Gu, Zhenfeng; Yin, Zhouping

doi:10.1109/tii.2023.3252405

Cited by 9 publications

(1 citation statement)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this paper, deep learning techniques based on CRNN model for Multivariate Time Series Prediction, combined with a method based on DAE to learn noise characteristics, have been exploited to generate synthetic sensor data with real world-like noise characteristics, incorporating measurements from sensors and heterogeneous information with the aim of generating synthetic data that can be used to augment real-world data. Specifically, we focused on two sensor classes: UWBs which are used in robotics for localization and SLAM tasks as widely employed in [78], [79], [80], [81], and [82] and UHF-RFID that have been used in autonomous agents for localization [83] and SLAM tasks as in [84] and [85].…”

Section: A Related Workmentioning

confidence: 99%

Synthetic Sensor Measurement Generation With Noise Learning and Multi-Modal Information

Romanelli,

Martinelli

2023

IEEE Access

View full text Add to dashboard Cite

Deep learning has transformed data generation, particularly in creating synthetic sensor data. This capability is invaluable in fields like autonomous driving, robotics, and computer science. To achieve this, we train models using real data, enabling them to replicate sensor data closely. These models introduce variations and noise, enhancing diversity and realism. Prominent techniques, including generative adversarial networks (GANs), variational autoencoders (VAEs), and recurrent neural networks (RNNs), excel in generating synthetic sensor data. Our paper focuses on Autoregressive Convolutional Recurrent Neural Networks (CRNN) for Multivariate Time Series Prediction. We incorporate Denoising Autoencoders (DAE) to mimic real-world noise characteristics. Our model is trained and validated using Ultra Wide Band (UWB) and Ultra High-Frequency Radio Frequency Identification (UHF-RFID) sensor data. It integrates sensor measurements and diverse information sources to produce synthetic data complementing realworld data. While demonstrated with UHF-RFID and UWB sensors, these techniques extend to industrial automation, healthcare and environmental monitoring. While our methodology exhibits broad potential, we present practical demonstrations with UHF-RFID and UWB sensors. Our deep neural network model allows researchers to construct datasets for algorithm validation, eliminating the need for costly and time-consuming data collection.

show abstract