Light Detection And Ranging Research Articles

A multi-sensor fusion SLAM algorithm for indoor aerial robots

Nowdays, existing Light Detection and Ranging (LiDAR) based unknown-space-exploration Simultaneous Localization and Mapping (SLAM) methods have the problems of unstable mapping in feature-degraded environments and poor estimation accuracy in the vertical direction. To solve these problems, this study presents a multi-sensor fusion SLAM algorithm for indoor aerial robots as well as Unmanned Aerial Vehicles (UAVs). The sensors used in the multi-sensor fusion algorithm include optical flow sensor, barometer, Inertial Measurement Unit (IMU), and LiDAR. The proposed algorithm is named LiDAR-IMU-optical flow odometry fusion algorithm (Lioo). We derived and established a tightly coupled multi-sensor fusion framework based on factor graph optimisation. The IMU-optical flow pre-integration factor is proposed, and the barometer factor to the algorithm is added innovatively. Engineering applications are also realised in this paper. Flight experiments demonstrated that the proposed algorithm achieves more precise mapping and localisation of UAVs in indoor environments.

Possibility of 3D Model Collections with LiDAR: Simply, Easily, and Inexpensively in Paleontology

Until several years ago, it was difficult for the general public to acquire 3D data. Recently, portable devices have been equipped with LiDAR (Light Detection And Ranging). LiDAR is a method for determining ranges by targeting an object or a surface with a laser and measuring the time for the reflected light to return to the receiver. It is easy to acquire 3D data by LiDAR at low cost, so making 3D digital specimens and using 3D databases have become more popular. This study aims to to investigate the possibility of creating 3D digital model collections in paleontology using a simple, easy and inexpensive approach that applies portable devices equipped with LiDAR. Two Apple iPad Pro® (3rd and 4th generation) and an iPhone® 15 Pro equipped with LiDAR are used with the Scaniverse application. The scanning distance of Scaniverse is from 0.3 m to 5.0 m for 3D modelling. Targets of paleontological 3D models in our study, range from hand-sized specimens to exhibits several meters in size, exhibition rooms and outcrops, including a 5 cm ammonite and an ~3 m full-body replica of Naumann’s elephant, Palaeoloxodon naumanni. In this case study, it was difficult to make 3D models of specimens under 10 cm diameter. On the other hand, 3D modelling of specimens greater than 10 cm, and up to 5 m long produced reasonable results. These 3D model collections can be used as exhibits and for education by making 3D printings and virtual reality (VR) models. Because the LiDAR technology is now built into common devices, problems may arise because anyone could make and share 3D models simply, easily, and inexpensively. Several issues of ownership, such as copyright, could arise from creating these models because digital data policies are not uniform or codified between museums and institutions. In the future, rules to govern these collections internationally need to be developed.

Retrieval performance of mangrove tree heights using multiple machine learning regression models and UAV-LiDAR point clouds

ABSTRACT Mangroves are vital coastal ecosystems that provide crucial links between land and sea. Tree height is a key indicator for assessing mangroves’ health status. Currently, there are still numerous challenges in estimating mangrove tree height. In this study, multiple deep learning and shallow machine learning regression models were developed to accurately estimate mangrove tree height using multi-dimensional Light Detection and Ranging (LiDAR) point clouds and their derivatives. We constructed a novel CNN_RepMLP model for mangrove tree height mapping. We also further verified the applicability of different types of regression models in estimating mangrove tree heights, and explored the influence of different LiDAR-derived features on the inversion accuracy for tree heights. The results indicated the following: (1) The CNN_RepMLP displayed satisfactory performance in the inversion of mangrove tree heights, and exhibited better robustness and generalization ability than the convolutional neural network (CNN) model. (2) Among the different LiDAR-derived feature combinations, combining height variables with intensity variables can not only mitigate the negative impact of height variables on inversion models, but also enhance the mangrove tree height inversion accuracy. (3) The ensemble learning framework with ExtraTrees as the meta-model can make better use of the differences and complementarities between different single base models, and exhibited better accuracy in estimating the height of mangrove trees compared with other ensemble learning models. (4) Multiple machine learning models based on multi-dimensional UAV-LiDAR point-cloud-derived features are suitable for the inversion of mangrove tree height. The CNN_RepMLP model outperformed the CNN and stacking ensemble learning models and had more detailed differentiation in terms of mangrove tree height. Its prediction results can more realistically reflect the spatial differentiation characteristics of mangrove tree height.

3D modelling method and application to a digital campus by fusing point cloud data and image data

ObjectiveThe use of single-source data for real-world 3D modelling currently faces problems such as deformation, pulling and fuzzy texture at the bottom of buildings in some feature models because of the lack of images. Moreover, LIDAR generates a huge amount of data, and the massive raw data processing and point cloud parsing puts high demands on the hardware arithmetic and algorithms. Aiming at the deficiencies and defects of the two data sources of inclined photogrammetry and airborne laser point cloud in the construction of high-quality and high-precision city-level 3D models. Methodsthis study uses a university library building as an example and proposes the main technical process and method of modelling after fusing the point cloud data acquired by inclined photogrammetry and 3D laser scanning technology. This is accomplished in the reconstruction stage of multi-source data fusion through data spatial alignment, coordinate system unification and data spatial integration. At the stage of multi-source data fusion and reconstruction, through data spatial alignment, coordinate system unification, point cloud coarse alignment and the iterative closest point (ICP) algorithm, a realistic 3D model of a building is constructed to verify the effectiveness of the modelling method. ResultsThe method can effectively improve the accuracy of the real-life 3D model, repair the deficiencies in the model and optimise the details of the model. It can also significantly improve the fineness of the tilt photography model and perfectly present the geometric and texture information of the building, making it a superior method for fine 3D reconstruction. ConclusionThis 3D reconstruction method of buildings, which integrates low-altitude inclined photogrammetry and airborne light detection and ranging (LiDAR), has high positional accuracy and can provide new methods and new ideas for the construction of digital campuses as well as for other engineering applications.

Seeing is believing: An Augmented Reality application for Palaeolithic rock art

By developing new recording methodologies, current rock art studies generate a large amount of graphic information about sites (tracings, photographs, three-dimensional reproductions) providing visibility of this fragile and little-known heritage, whose accessibility is often difficult or impossible for the general public. In addition, many rock art depictions are challenging to observe, due to the very nature of the artistic entities (fine engravings or faded paintings in karst environments or open-air sites with poor or changing light conditions), or to conservation problems derived from natural factors such as erosion and geological and biological processes, as well as from anthropic factors. These conditions make rock art depictions nearly indistinguishable in many places and on many objects today, except for experts. This difficulty of accessing and visualising rock art heritage, located in fragile environments and often challenging places such as caves or difficult-to-reach open-air sites, makes the information and knowledge generated by investigation of this heritage asset difficult to transfer to society in general, which is frequently unaware of the priceless value of this heritage. The present study proposes generating several mechanisms to transfer the results of research, restitution and documentation of rock art to society in general. An AR (Augmented Reality) application has been developed using LiDAR (Light Detection and Ranging) technology to address current challenges in implementing AR technologies in low-light environments. So far, this app has been developed in a Proof-of-Concept project at Spanish archaeological sites such as Hornos de la Peña (Cantabria), Domingo García (Segovia) and La Salud (Salamanca). This application will be particularly interesting for sites currently visited with or without a guide, allowing user interactivity and real-time reconstruction, for example, of the visibility of graphic motifs.

Multi-level interactive fusion network based on adversarial learning for fusion classification of hyperspectral and LiDAR data

With the growing awareness of the limitations of unimodal data, research on joint classification using multimodal remote sensing data has garnered increasing attention. However, existing methods still have certain deficiencies in dealing with feature extraction and feature fusion of multimodal remote sensing data. In order to obtain more accurate classification results, it is crucial to fully exploit and utilize the complementary information in multimodal remote sensing data. In this paper, a multi-level interactive fusion network based on adversarial learning (MLIF-AL) is proposed for hyperspectral image (HSI) and light detection and ranging (LiDAR) data fusion classification. First, the feature extraction part based on generative adversarial network (GAN) utilizes adversarial training between the generator and the discriminator to guide the network to learn more discriminative and distinguished feature representations. Meanwhile, a cross-modal interactive information extraction (CMIIE) module is also designed for the generator encoding part to promote the interaction between multimodal data from a global perspective and realize the full utilization of multimodal complementary information. In addition, in the multi-level feature fusion classification (MLFFC) part, the complementarities between the features learned by the neural networks at each layer are comprehensively utilized to obtain higher-level multimodal fusion semantic information. Finally, the adversarial loss and classification loss are combined for network training. Extensive experiments on three commonly used HSI and LiDAR datasets and comparisons with state-of-the-art methods demonstrate the effectiveness and superiority of the model.

Hydraulic Risk Assessment on Historic Masonry Bridges Using Hydraulic Open-Source Software and Geomatics Techniques: A Case Study of the “Hannibal Bridge”, Italy

This paper investigates the impact of flood-induced hydrodynamic forces and high discharge on the masonry arch “Hannibal Bridge” (called “Ponte di Annibale” in Italy) using the Hydraulic Engineering Center’s River Analysis Simulation (HEC-RAS) v6.5.0. hydraulic numerical method, incorporating Unmanned Aerial Vehicle (UAV) photogrammetry and aerial Light Detection and Ranging (LIDAR) data for visual analysis. The research highlights the highly transient behavior of fast flood flows, particularly when carrying debris, and their effect on bridge superstructures. Utilizing a Digital Elevation Model to extract cross-sectional and elevation data, the research examined 23 profiles over 800 m of the river. The results indicate that the maximum allowable water depth in front of the bridge is 4.73 m, with a Manning’s coefficient of 0.03 and a longitudinal slope of 9 m per kilometer. Therefore, a novel method to identify the risks through HEC-RAS modeling significantly improves the conservation of masonry bridges by providing precise topographical and hydrological data for accurate simulations. Moreover, the detailed information obtained from LIDAR and UAV photogrammetry about the bridge’s materials and structures can be incorporated into the conservation models. This comprehensive approach ensures that preservation efforts are not only addressing the immediate hydrodynamic threats but are also informed by a thorough understanding of the bridge’s structural and material conditions. Understanding rating curves is essential for water management and flood forecasting, with the study confirming a Manning roughness coefficient of 0.03 as suitable for smooth open-channel flows and emphasizing the importance of geomorphological conditions in hydraulic simulation.

On the validation of a modelling tool for Vehicle Integrated PhotoVoltaics: Reflected irradiance in urban environments

Vehicle Integrated PhotoVoltaics (VIPV) technology holds immense promise for extending the operational range of Electric Vehicles by integrating Photovoltaic cells onto vehicle surfaces. However, VIPV faces unique challenges coming from the dynamic nature of vehicular movement and the complex urban environment. This paper presents a comprehensive methodology for the analysis of VIPV systems, addressing these challenges. Light Detection and Ranging (LiDAR) technology provides a high-resolution point cloud to characterize urban topography and enables to analyse of the impact of urban objects on the different components of the irradiance. The proposed methodology is validated through a measurement campaign, demonstrating the accuracy of the proposed modelling. By bridging empirical data collection with computational modelling, this research provides a robust foundation for analysing VIPV systems in urban environments.

Research on the Method for Recognizing Bulk Grain-Loading Status Based on LiDAR.

Grain is a common bulk cargo. To ensure optimal utilization of transportation space and prevent overflow accidents, it is necessary to observe the grain's shape and determine the loading status during the loading process. Traditional methods often rely on manual judgment, which results in high labor intensity, poor safety, and low loading efficiency. Therefore, this paper proposes a method for recognizing the bulk grain-loading status based on Light Detection and Ranging (LiDAR). This method uses LiDAR to obtain point cloud data and constructs a deep learning network to perform target recognition and component segmentation on loading vehicles, extract vehicle positions and grain shapes, and recognize and make known the bulk grain-loading status. Based on the measured point cloud data of bulk grain loading, in the point cloud-classification task, the overall accuracy is 97.9% and the mean accuracy is 98.1%. In the vehicle component-segmentation task, the overall accuracy is 99.1% and the Mean Intersection over Union is 96.6%. The results indicate that the method has reliable performance in the research tasks of extracting vehicle positions, detecting grain shapes, and recognizing loading status.

Detailed height mapping of trees and buildings (HiTAB) in Chicago and its implications to urban climate studies

Abstract The vertical dimensions of urban morphology, specifically the heights of trees and buildings, exert significant influence on wind flow fields in urban street canyons and the thermal environment of the urban fabric, subsequently affecting the microclimate, noise levels, and air quality. Despite their importance, these critical attributes are less commonly available and rarely utilized in urban climate models compared to planar land use and land cover data. In this study, we explicitly mapped the height of trees and buildings (HiTAB) across the city of Chicago at 1 m spatial resolution using a data fusion approach. This approach integrates high-precision light detection and ranging (LiDAR) cloud point data, building footprint inventory, and multi-band satellite images. Specifically, the digital terrain and surface models were first created from the LiDAR dataset to calculate the height of surface objects, while the rest of the datasets were used to delineate trees and buildings. We validated the derived height information against the existing building database in downtown Chicago and the Meter-scale Urban Land Cover map from the Environmental Protection Agency, respectively. The co-investigation on trees and building heights offers a valuable initiative in the effort to inform urban land surface parameterizations using real-world data. Given their high spatial resolution, the height maps can be adopted in physical-based and data-driven urban models to achieve higher resolution and accuracy while lowering uncertainties. Moreover, our method can be extended to other urban regions, benefiting from the growing availability of high-resolution urban informatics globally. Collectively, these datasets can substantially contribute to future studies on hyper-local weather dynamics, urban heterogeneity, morphology, and planning, providing a more comprehensive understanding of urban environments.

QGIS‐based support tools to simplify the complex design challenges of subsurface drainage systems

AbstractWe developed a comprehensive set of tools to simplify the design and layout of subsurface drainage systems. These tools, which utilize publicly available LiDAR (light detection and ranging) datasets, may also be used for other applications. The tool streamline operations include harmonizing geographic information system (GIS) layers into a single State Plane or Universal Transverse Mercator Coordinate System, clipping and thinning LiDAR data within a defined boundary, generating drainage networks with contour lines and laylines for surface flow visualization, identifying depressions and creating gridlines with user‐specified angles and spacings. To showcase their use, an illustrative example using a diverse dataset from the University of Illinois’ South Farm is provided. These user‐friendly tools, optimized for compatibility with all versions of QGIS3 (quantum geographic information system 3), are freely accessible for download or installation from the official QGIS plugin repository. A user manual with step‐by‐step instructions is available online on the Illinois Drainage Guide website.

Implementation of a Small-Sized Mobile Robot with Road Detection, Sign Recognition, and Obstacle Avoidance

In this study, under the limited volume of 18 cm × 18 cm × 21 cm, a small-sized mobile robot is designed and implemented. It consists of a CPU, a GPU, a 2D LiDAR (Light Detection And Ranging), and two fisheye cameras to let the robot have good computing processing and graphics processing capabilities. In addition, three functions of road detection, sign recognition, and obstacle avoidance are implemented on this small-sized robot. For road detection, we divide the captured image into four areas and use Intel NUC to perform road detection calculations. The proposed method can significantly reduce the system load and also has a high processing speed of 25 frames per second (fps). For sign recognition, we use the YOLOv4-tiny model and a data augmentation strategy to significantly improve the computing performance of this model. From the experimental results, it can be seen that the mean Average Precision (mAP) of the used model has increased by 52.14%. For obstacle avoidance, a 2D LiDAR-based method with a distance-based filtering mechanism is proposed. The distance-based filtering mechanism is proposed to filter important data points and assign appropriate weights, which can effectively reduce the computational complexity and improve the robot’s response speed to avoid obstacles. Some results and actual experiments illustrate that the proposed methods for these three functions can be effectively completed in the implemented small-sized robot.

Modelling and Analysis of Vector and Vector Vortex Beams Reflection for Optical Sensing

Light Detection and Ranging (LiDAR) sensors can precisely determine object distances using the pulsed time of flight (TOF) or amplitude-modulated continuous wave (AMCW) TOF methods and velocity using the frequency-modulated continuous wave (FMCW) approach. In this paper, we focus on modelling and analysing the reflection of vector beams (VBs) and vector vortex beams (VVBs) for optical sensing in LiDAR applications. Unlike traditional TOF and FMCW methods, this novel approach uses VBs and VVBs as detection signals to measure the orientation of reflecting surfaces. A key component of this sensing scheme is understanding the relationship between the characteristics of the reflected optical fields and the orientation of the reflecting surface. To this end, we develop a computational model for the reflection of VBs and VVBs. This model allows us to investigate critical aspects of the reflected field, such as intensity distribution, intensity centroid offset, reflectance, and the variation of the intensity range measured along the azimuthal direction. By thoroughly analysing these characteristics, we aim to enhance the functionality of LiDAR sensors in detecting the orientation of reflecting surfaces.

Assessing allergy risk from ornamental trees in a city: Integrating open access remote sensing data with pollen measurements

Platanus sp. pl. (plane trees) are common ornamental tree in Poland that produces a large amount of wind-transported pollen, which contains proteins that induce allergy symptoms. Allergy sufferers can limit their contact with pollen by avoiding places with high pollen concentrations, which are restricted mainly to areas close to plane trees. Their location is thus important, but creating a detailed street tree inventory is expensive and time-consuming. However, high-resolution remote sensing data provide an opportunity to detect the location of specific plants. But acquiring high-resolution spatial data of good quality also incurs costs and requires regular updates. Therefore, this study explored the potential of using open access remote sensing data to detect plane trees in the highly urbanized environment of Poznań (western Poland). Airborne light detection and ranging (LiDAR) was used to detect training treetops, which were subsequently marked as young plane trees, mature plane trees, other trees or artefacts. Spectral and spatial variables were extracted from circular buffers (r = 1 m) around the treetops to minimize the influence of shadows and crown overlap. A random forest machine learning algorithm was applied to assess the importance of variables and classify the treetops within a radius of 6.2 km around the functioning pollen monitoring station. The model performed well during 10-fold cross-validation (overall accuracy ≈ 92%). The predicted Platanus sp. pl. locations, aggregated according to 16 wind directions, were significantly correlated with the hourly pollen concentrations. Based on the correlation values, we established a threshold of prediction confidence, which allowed us to reduce the fraction of false-positive predictions. We proposed the spatially continuous index of airborne pollen exposure probability, which can be useful for allergy sufferers. The results showed that open-access geodata in Poland can be applied to recognize major local sources of plane pollen.

Design and testing of an intelligent variable-rate spraying system based on LiDAR-guided application

Traditional plant protection machinery cannot apply pesticides according to the actual growth demand of crops, which causes a lot of pesticide waste and serious environmental pollution. Based on this, an intelligent variable-rate application system based on 3D LiDAR (light detection and ranging) scanning and PWM (pulse-width modulation) control for maize canopy was developed in this paper. After completing the performance testing, the system was mounted on a lightweight high-clearance self-propelled sprayer by remote controlling with an operating width of 6.0 m, and a ground gap of 2.0 m. And the actual operational performance of the sprayer showed that the prediction error of LiDAR for plant height was within 10%, which illustrated the accuracy of the detection system and the reliability of the variable application based on the plant height change. And the variable-rate system could adjust the spray volume in real-time according to the plant height change. The penetration rate of the upper, middle, and bottom layers maintained a relatively stable state (upper > middle > bottom) at variable-rate application and could reach the normal application level. This study can provide new ideas for efficient pests and diseases control of high stalk crops, especially suitable for small fields, and gives a reference for the structural design of new precise variable-rate application machinery and performance optimization of efficient application technology.

Room-temperature DBR-free VCSEL operation on an InGaN/GaN thin-film platform with a monolithic surface grating.

Vertical-cavity surface-emitting lasers (VCSELs) are pivotal in various applications ranging from data communication to sensing technologies. This study introduces a VCSEL design featuring a monolithic top surface high contrast grating (HCG) reflector on a thin-film substrate, aimed at improving lasing performance while reducing fabrication costs by omitting the use of distributed Bragg reflectors (DBRs). We fabricated the proposed VCSEL with the surface grating and characterized its performance through micro-photoluminescence measurements. The laser demonstrated room-temperature lasing at 436.2 nm with a Q factor of 4600 and a lasing threshold of 5.5 kW/cm2 under optical pumping. The implementation of the surface grating reflector was instrumental in facilitating vertical lasing, significantly improving surface reflectivity compared to conventional flat GaN/air interfaces. This innovative design holds significant promise for the development of cost-effective, DBR-free VCSELs, with potential applications extending to photonic integrated circuits and light detection and ranging (LiDAR) systems.

Towards an ultrafast 3D imaging scanning LiDAR system: a review

Light detection and ranging (LiDAR), as a hot imaging technology in both industry and academia, has undergone rapid innovation and evolution. The current mainstream direction is towards system miniaturization and integration. There are many metrics that can be used to evaluate the performance of a LiDAR system, such as lateral resolution, ranging accuracy, stability, size, and price. Until recently, with the continuous enrichment of LiDAR application scenarios, the pursuit of imaging speed has attracted tremendous research interest. Particularly, for autonomous vehicles running on motorways or industrial automation applications, the imaging speed of LiDAR systems is a critical bottleneck. In this review, we will focus on discussing the upper speed limit of the LiDAR system. Based on the working mechanism, the limitation of optical parts on the maximum imaging speed is analyzed. The beam scanner has the greatest impact on imaging speed. We provide the working principle of current popular beam scanners used in LiDAR systems and summarize the main constraints on the scanning speed. Especially, we highlight the spectral scanning LiDAR as a new paradigm of ultrafast LiDAR. Additionally, to further improve the imaging speed, we then review the parallel detection methods, which include multiple-detector schemes and multiplexing technologies. Furthermore, we summarize the LiDAR systems with the fastest point acquisition rate reported nowadays. In the outlook, we address the current technical challenges for ultrafast LiDAR systems from different aspects and give a brief analysis of the feasibility of different approaches.

Enhancing Autonomous Truck Navigation with Ultra-Wideband Technology in Industrial Environments.

The integration of autonomous vehicles in industrial settings necessitates advanced positioning and navigation systems to ensure operational safety and efficiency. This study rigorously evaluates the application of Ultra-Wideband (UWB) technology in autonomous industrial trucks and compares its effectiveness with conventional systems such as Light Detection and Ranging (LiDAR), Global Positioning System (GPS), and cameras. Through comprehensive experiments conducted in a real factory environment, this study meticulously assesses the accuracy and reliability of UWB technology across various reference distances and under diverse environmental conditions. The findings reveal that UWB technology consistently achieves positioning accuracy within 0.2 cm 99% of the time, significantly surpassing the 10 cm and 5 cm accuracies of GPS and LiDAR, respectively. The exceptional performance of UWB, especially in environments afflicted by high metallic interference and non-line-of-sight conditions-where GPS and LiDAR's efficacy decreased by 40% and 25%, respectively-highlights its potential to revolutionize the operational capabilities of autonomous trucks in industrial applications. This study underscores the robustness of UWB in maintaining high accuracy even in adverse conditions and illustrates its low power consumption and efficiency in multi-user scenarios without signal interference. This study not only confirms the superior capabilities of UWB technology but also contributes to the broader field of autonomous vehicle technology by highlighting the practical benefits and integration potential of UWB systems in complex and dynamic environments.

A 128 × 128 SPAD LiDAR sensor with column-parallel 25 ps resolution TA-ADCs

This paper presents a design of single photon avalanche diode (SPAD) light detection and ranging (LiDAR) sensor with 128 × 128 pixels and 128 column-parallel time-to-analog-merged-analog-to-digital converts (TA-ADCs). Unlike the conventional TAC-based SPAD LiDAR sensor, in which the TAC and ADC are separately implemented, we propose to merge the TAC and ADC by sharing their capacitors, thus avoiding the analog readout noise of TAC’s output buffer, improving the conversion rate, and reducing chip area. The reverse start-stop logic is employed to reduce the power of the TA-ADC. Fabricated in a 180 nm CMOS process, our prototype sensor exhibits a timing resolution of 25 ps, a DNL of +0.30/−0.77 LSB, an INL of +1.41/−2.20 LSB, and a total power consumption of 190 mW. A flash LiDAR system based on this sensor demonstrates the function of 2D/3D imaging with 128 × 128 resolution, 25 kHz inter-frame rate, and sub-centimeter ranging precision.

Cross-Scene Joint Classification of Multisource Data With Multilevel Domain Adaption Network.

Domain adaption (DA) is a challenging task that integrates knowledge from source domain (SD) to perform data analysis for target domain. Most of the existing DA approaches only focus on single-source-single-target setting. In contrast, multisource (MS) data collaborative utilization has been extensively used in various applications, while how to integrate DA with MS collaboration still faces great challenges. In this article, we propose a multilevel DA network (MDA-NET) for promoting information collaboration and cross-scene (CS) classification based on hyperspectral image (HSI) and light detection and ranging (LiDAR) data. In this framework, modality-related adapters are built, and then a mutual-aid classifier is used to aggregate all the discriminative information captured from different modalities for boosting CS classification performance. Experimental results on two cross-domain datasets show that the proposed method consistently provides better performance than other state-of-the-art DA approaches.