Vision-and-language navigation (VLN) is a challenging task that requires an agent to navigate in real-world environments by understanding natural language instructions and visual information received in real time. Prior works have implemented VLN tasks on continuous environments or physical robots, all of which use a fixed-camera configuration due to the limitations of datasets, such as 1.5-m height, 90° horizontal field of view (HFOV), and so on. However, real-life robots with different purposes have multiple camera configurations, and the huge gap in visual information makes it difficult to directly transfer the learned navigation skills between various robots. In this brief, we propose a visual perception generalization strategy based on meta-learning, which enables the agent to fast adapt to a new camera configuration. In the training phase, we first locate the generalization problem to the visual perception module and then compare two meta-learning algorithms for better generalization in seen and unseen environments. One of them uses the model-agnostic meta-learning (MAML) algorithm that requires few-shot adaptation, and the other refers to a metric-based meta-learning method with a feature-wise affine transformation (AT) layer. The experimental results on the VLN-CE dataset demonstrate that our strategy successfully adapts the learned navigation skills to new camera configurations, and the two algorithms show their advantages in seen and unseen environments respectively.

Prior works in vision-and-language navigation (VLN) focus on using long short-term memory (LSTM) to carry the flow of information on either the navigation model (navigator) or the instruction generating model (speaker).The outstanding capability of LSTM to process intermodal interactions has been widely verified; however, LSTM neglects the intramodel interactions, leading to negative effect on either navigator or speaker. The performance of attention-based Transformer is satisfactory in sequence-to-sequence translation domains, but Transformer structure implemented directly in VLN has yet been satisfied. In this article, we propose novel Transformer-based multimodal frameworks for the navigator and speaker, respectively. In our frameworks, the multihead self-attention with the residual connection is used to carry the information flow. Specially, we set a switch to prevent them from directly entering the information flow in our navigator framework. In experiments, we verify the effectiveness of our proposed approach, and show significant performance advantages over the baselines.

Micro electro mechanical system (MEMS) gyroscopes and magnetometers are usually integrated into a sensor module or chip and widely used in a variety of applications. In existing integrated gyroscope and magnetometer calibration methods, rotation in all possible orientations is a necessary condition for a good calibration result. However, rotation around two or more axes is difficult to attain, as it is limited by the range of movement of vehicles such as cars, ships, or planes. To solve this problem, this paper proposes an integrated magnetometer and gyroscope calibration method with level rotation. The proposed method presents a redefined magnetometer output model using level attitude. New gyroscope and magnetometer calibration models are then deduced. In addition, a simplified cubature Kalman filter (CKF) is established to estimate calibration parameters. This method possesses important value for application in actual systems, as it only needs level rotation for real-time calibration of gyroscopes and magnetometers. Theoretical analysis and test results verify the validity and feasibility of this method.

The integration of magnetometers and Inertial Navigation Systems (INS) is widely used in low-cost navigation systems. However, even if the system has been calibrated, random magnetic disturbances still appear in practical applications, which lead to large heading errors. To solve this problem, an adaptive anti-disturbance method to overcome random magnetic disturbance is proposed. First, disturbances are classified and analysed in detail based on actual road vehicle driving data. Then an Adaptive Robust Extend Kalman Filter (AREKF) is designed to resist sudden disturbances. However, an AREKF may accumulate errors slowly when a long-term disturbance exists. Considering this situation, this paper proposes that AREKF is used to maintain accuracy in the early stages, at the same time as the magnetometer is quickly calibrated with a Kalman filter. Then, the new magnetometer parameters are put into the AREKF to suppress long-term disturbances. Finally, cascading these two modules, not only the sudden disturbance can be overcome, but the situation of long-term disturbances can be suppressed. The results of simulation and an actual driving test show that the proposed method can effectively overcome random magnetic disturbances in both the short and long term.