Trainable Quaternion Extended Kalman Filter with Multi-Head Attention for Dead Reckoning in Autonomous Ground Vehicles.

Extended Kalman filter (EKF) is one of the most widely used Bayesian estimation methods in the optimal control area. Recent works on mobile robot control and transportation systems have applied various EKF methods, especially for localization. However, it is difficult to obtain adequate and reliable process-noise and measurement-noise models due to the complex and dynamic surrounding environments and sensor uncertainty. Generally, the default noise values of the sensors are provided by the manufacturer, but the values may frequently change depending on the environment. Thus, this paper mainly focuses on designing a highly accurate trainable EKF-based localization framework using inertial measurement units (IMUs) for the autonomous ground vehicle (AGV) with dead reckoning, with the goal of fusing it with a laser imaging, detection, and ranging (LiDAR) sensor-based simultaneous localization and mapping (SLAM) estimation for enhancing the performance. Convolution neural networks (CNNs), backward propagation algorithms, and gradient descent methods are implemented in the system to optimize the parameters in our framework. Furthermore, we develop a unique cost function for training the models to improve EKF accuracy. The proposed work is general and applicable to diverse IMU-aided robot localization models.

DXM-TransFuse U-net: Dual cross-modal transformer fusion U-net for automated nerve identification.

Accurate nerve identification is critical during surgical procedures to prevent damage to nerve tissues. Nerve injury can cause long-term adverse effects for patients, as well as financial overburden. Birefringence imaging is a noninvasive technique derived from polarized images that have successfully identified nerves that can assist during intraoperative surgery. Furthermore, birefringence images can be processed under 20 ms with a GPGPU implementation, making it a viable image modality option for real-time processing. In this study, we first comprehensively investigate the usage of birefringence images combined with deep learning, which can automatically detect nerves with gains upwards of 14% over its color image-based (RGB) counterparts on the F2 score. Additionally, we develop a deep learning network framework using the U-Net architecture with a Transformer based fusion module at the bottleneck that leverages both birefringence and RGB modalities. The dual-modality framework achieves 76.12 on the F2 score, a gain of 19.6 % over single-modality networks using only RGB images. By leveraging and extracting the feature maps of each modality independently and using each modality's information for cross-modal interactions, we aim to provide a solution that would further increase the effectiveness of imaging systems for enabling noninvasive intraoperative nerve identification.