Monocular visual inertial odometry (VIO) has facilitated a wide range of real-time motion tracking applications, thanks to the small size of the sensor suite and low power consumption. To successfully bootstrap VIO algorithms, the initialization module is extremely important. Most initialization methods rely on the reconstruction of 3D visual point clouds. These methods suffer from high computational cost as state vector contains both motion states and 3D feature points. To address this issue, some researchers recently proposed a structureless initialization method, which can solve the initial state without recovering 3D structure. However, this method potentially compromises performance due to the decoupled estimation of rotation and translation, as well as linear constraints. To improve its accuracy, we propose novel structureless visual-inertial bundle adjustment to further refine previous structureless solution. Extensive experiments on real-world datasets show our method significantly improves the VIO initialization accuracy, while maintaining real-time performance.

Concept drift remains a critical challenge in non-stationary industrial fault diagnosis processes, where evolving operational conditions induce continuous distributional shifts and hinder model generalization. While existing continual learning approaches primarily focus on mitigating forgetting, they often lack explicit mechanisms to detect and quantify drift in a timely manner. To address this, this paper proposes a novel multi-scale feature-probability consistency-based drift detection (MFPC-DCD) framework. It comprises three core components: (1) a feature-level divergence module that leverages Maximum Mean Discrepancy (MMD) to measure latent representation shifts; (2) a probability-level divergence module utilizing Jensen–Shannon divergence to quantify predictive inconsistencies; and (3) a multi-scale consistency fusion strategy that aggregates drift signals across short-, mid-, and long-term temporal windows, thereby enhancing robustness and temporal granularity. The MFPC-DCD method can explicitly detect domain shifts by jointly capturing structural and semantic distributional variations through multi-resolution analysis and then the resulting unified drift coefficient provides precise quantification of domain shifts and can be seamlessly integrated into downstream continual learning fault diagnosis for adaptive regularization. Experiments conducted on two fault diagnosis datasets — CWRU, and our proprietary DPS — demonstrate that MFPC-DCD achieves state-of-the-art drift detection performance and adaptability under diverse drift scenarios.

Online extrinsic calibration is crucial for building "power-on-and-go" moving platforms, like robots and AR devices. However, blindly performing online calibration for unobservable parameter may lead to unpredictable results. In the literature, extensive studies have been conducted on the extrinsic calibration between IMU and camera, from theory to practice. It is well-known that the observability of extrinsic parameter can be guaranteed under sufficient motion excitation. Furthermore, the impacts of degenerate motions are also investigated. Despite these successful analyses, we identify an issue with respect to the existing observability conclusion. This paper focuses on the observability investigation for straight line motion, which is a common-seen and fundamental degenerate motion in applications. We analytically prove that pure translational straight line motion can lead to the unobservability of the rotational extrinsic parameter between IMU and camera (at least one degree of freedom). By correcting the existing observability conclusion, our novel theoretical finding disseminates more precise principle to the research community and provides explainable calibration guideline for practitioners. Our analysis is validated by rigorous theory and experiments.

In robotics, motion capture systems have been widely used to measure the accuracy of localization algorithms. Moreover, this infrastructure can also be used for other computer vision tasks, such as the evaluation of Visual (-Inertial) SLAM dynamic initialization, multi-object tracking, or automatic annotation. Yet, to work optimally, these functionalities require having accurate and reliable spatial-temporal calibration parameters between the camera and the global pose sensor. In this study, we provide two novel solutions to estimate these calibration parameters. Firstly, we design an offline target-based method with high accuracy and consistency. Spatial-temporal parameters, camera intrinsic, and trajectory are optimized simultaneously. Then, we propose an online target-less method, eliminating the need for a calibration target and enabling the estimation of time-varying spatial-temporal parameters. Additionally, we perform detailed observability analysis for the target-less method. Our theoretical findings regarding observability are validated by simulation experiments and provide explainable guidelines for calibration. Finally, the accuracy and consistency of two proposed methods are evaluated with hand-held real-world datasets where traditional hand-eye calibration method do not work.

Accurate global localization is crucial for autonomous navigation and planning. To this end, various GPS-aided Visual-Inertial Odometry (GPS-VIO) fusion algorithms are proposed in the literature. This paper presents a novel GPS-VIO system that is able to significantly benefit from the online calibration of the rotational extrinsic parameter between the GPS reference frame and the VIO reference frame. The behind reason is this parameter is observable. This paper provides novel proof through nonlinear observability analysis. We also evaluate the proposed algorithm extensively on diverse platforms, including flying UAV and driving vehicle. The experimental results support the observability analysis and show increased localization accuracy in comparison to state-of-the-art (SOTA) tightly-coupled algorithms.

Deep learning for cell instance segmentation is a significant research direction in biomedical image analysis. The traditional supervised learning methods rely on pixel-wise annotation of object images to train the models, which is often accompanied by time-consuming and labor-intensive. Various modified segmentation methods, based on weakly supervised or semi-supervised learning, have been proposed to recognize cell regions by only using rough annotations of cell positions. However, it is still hard to achieve the fully unsupervised in most approaches that the utilization of few annotations for training is still inevitable. In this article, we propose an end-to-end unsupervised model that can segment individual cell regions on hematoxylin and eosin (H&E) stained slides without any annotation. Compared with weakly or semi-supervised methods, the input of our model is in the form of raw data without any identifiers and there is no need to generate pseudo-labelling during training. We demonstrated that the performance of our model is satisfactory and also has a great generalization ability on various validation sets compared with supervised models. The ablation experiment shows that our backbone has superior performance in capturing object edge and context information than pure CNN or transformer under our unsupervised method.

Fast and accurate road damage detection is essential for the automatization of road inspection. This paper describes our solution submitted to the Global Road Damage Detection Challenge of the 2020 IEEE International Conference on Big Data, for typical road damage detection in digital images based on deep learning. The recently proposed YOLOv4 is chosen as the baseline network, while the effects of data augmentation, transfer learning, Optimized Anchors, and their combination are evaluated. We propose a novel road damage data generation method based on a generative adversarial network, which can generate multi-class samples with a single model. The evaluation results demonstrate the effectiveness of different tricks and their combinations on the road damage detection task, which provides a reference for practical application. The code of our solution is available at https://github.com/ZhangXG001/RoadDamgeDetection.git.

This paper presents a novel Cable Climbing Robot CCRobot-III, which is the third version designed for bridge cable inspection tasks, aiming at surpassing previous versions in terms of climbing speed and payload capacity. Benefiting from Split-type Wire-driven design, CCRobot-III can climb along a 90-110mm diameter bridge cable in inchworm-like gait at a speed of up to 12m/min, and carrying more than 40kg payload at the same time. CCRobot-III consists of a climbing precursor and a main-body frame. The two parts are connected and driven by steel wires. The climbing precursor, acting as a mobile anchor, moves quickly on a bridge cable. The mainbody frame, acting as a mobile winch, carries payload and pulls itself to a certain position with steel wires. Both parts have one or two pairs of palm-based gripper, which is the key component for providing strong adhesion to support the robot climbing. Experimental results have shown that CCRobotIII possesses outstanding climbing performance, high payload capacity, and good adaptability to complex conditions of cable surface. Moreover, it has potential engineering applications on the cable-stayed bridge for fieldwork.