Effective Indoor Robot Localization by Stacked Bidirectional LSTM Using Beacon-Based Range Measurements

Lim, Hyungtae; Myung, Hyun

doi:10.1007/978-981-13-7780-8_12

Cited by 3 publications

(6 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Finally, our fusion-based method was quantitatively compared with both the conventional and deep learning-based methods, namely, Synthetic in (12), CF in (14), UKF that takes Synthetic as measurements, ResNet [14], and Bi-LSTM [15]. For learning-based methods, we set the shape of the first kernel in ResNet and Bi-LSTM to suit our task.…”

Section: Comparison With the State-of-the-art Methodsmentioning

confidence: 99%

“…Therefore, our proposed method is confirmed to be a more appropriate and robust method for synthetic airspeed estimation and can help UAVs protect themselves against sensor failures even during aggressive flight maneuvers. Besides, our method only takes 6.7 msec for inference in offline mode because it requires a few FLOPS compared with other deep learning models [14], [15].…”

Section: Comparison With the State-of-the-art Methodsmentioning

confidence: 99%

“…To resolve these problems, we propose a learning-based method with TCN for robust synthetic airspeed estimation to overcome the aforementioned limitations of the conventional approaches. TCN [18] has strong advantages over the typical variants of recurrent neural networks (RNNs), such as LSTM and the gated recurrent unit (GRU) [15], in that TCN avoids the exploding/vanishing gradient problem, which is a major issue of RNNs. In addition, TCN needs a small amount of memory when training [28].…”

Section: B Temporal Convolutional Networkmentioning

confidence: 99%

“…Moreover, in the current era of deep learning [14], numerous attempts have been made to implement neural networks for various regression tasks, e.g., [15]- [18]. In particular, some researchers have adopted these machine learningbased approaches to aerospace applications.…”

Section: Introductionmentioning

confidence: 99%

“…2) The proposed deep learning model for synthetic airspeed estimation is shown to outperform conventional methods with small mean error and variance values. In addition, the temporal convolutional network (TCN) is shown to provide the most promising results compared with other deep learning models (i.e., ResNet [14] and bidirectional LSTM [15]) for synthetic airspeed estimation.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Deep Learning-Aided Synthetic Airspeed Estimation of UAVs for Analytical Redundancy With a Temporal Convolutional Network

Lim

Ryu

Rhudy

et al. 2022

IEEE Robot. Autom. Lett.

Self Cite

View full text Add to dashboard Cite

A synthetic air data system (SADS) is an analytical redundancy technique that is crucial for unmanned aerial vehicles (UAVs) and is used as a backup system during air data sensor failures. Unfortunately, the existing state-of-theart approaches for SADS require GPS signals or high-fidelity dynamic UAV models. To address this problem, a novel synthetic airspeed estimation method that leverages deep learning and an unscented Kalman filter (UKF) for analytical redundancy is proposed. Our novel fusion-based method only requires an inertial measurement unit (IMU), elevator control input, and airflow angles while GPS, lift/drag coefficients, and complex aircraft dynamic models are not required. Additionally, we demonstrate that our proposed temporal convolutional network (TCN) is a more efficient model for airspeed estimation than the renowned models, such as ResNet or bidirectional long short-term memory (LSTM). Our deep learning-aided UKF was experimentally verified on long-duration real flight data and has promising performance compared with the state-of-theart methods. In particular, it is confirmed that our proposed method robustly estimates the airspeed even under dynamic flight conditions where the performance of conventional methods is degraded.

show abstract

Section: Comparison With the State-of-the-art Methodsmentioning

confidence: 99%

Section: Comparison With the State-of-the-art Methodsmentioning

confidence: 99%

Section: B Temporal Convolutional Networkmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Deep Learning-Aided Synthetic Airspeed Estimation of UAVs for Analytical Redundancy With a Temporal Convolutional Network

Lim

Ryu

Rhudy

et al. 2022

IEEE Robot. Autom. Lett.

Self Cite

View full text Add to dashboard Cite

show abstract

Short-Term Traffic Speed Prediction Using Hybrid LSTM-SVR Model

Quach¹,

Ren

Tran

et al. 2023

Lecture Notes in Networks and Systems

View full text Add to dashboard Cite

A 3D Range-Only SLAM Algorithm Based on Improved Derivative UKF

et al. 2022

View full text Add to dashboard Cite

In this study, we constructed a 3D range-only (RO) localization algorithm based on improved unscented Kalman filtering (UKF). The algorithm can determine the location of unknown UWB nodes in a 3D environment through a moving node with low computational complexity, which can help agents to accurately identify feature points in 3D SLAM based only on the range. Specifically, we established an original UKF framework based the 3D RO localization algorithm, and developed a derivative UKF framework to reduce the computational complexity of the algorithm. We used singular value decomposition to compensate for the robustness of the algorithm. Next, we performed a theoretical analysis to show that our method reduces the computational burden without reducing the stability or accuracy of the system. Finally, we conducted numerical simulations and physical experiments to show the effectiveness of the developed 3D RO localization algorithm.

show abstract

Effective Indoor Robot Localization by Stacked Bidirectional LSTM Using Beacon-Based Range Measurements

Cited by 3 publications

References 23 publications

Deep Learning-Aided Synthetic Airspeed Estimation of UAVs for Analytical Redundancy With a Temporal Convolutional Network

Deep Learning-Aided Synthetic Airspeed Estimation of UAVs for Analytical Redundancy With a Temporal Convolutional Network

Short-Term Traffic Speed Prediction Using Hybrid LSTM-SVR Model

A 3D Range-Only SLAM Algorithm Based on Improved Derivative UKF

Contact Info

Product

Resources

About