This paper presents a novel finite-time adaptive fuzzy asymptotic tracking control framework specifically designed to meet the high-precision trajectory tracking requirements of quadrotor unmanned aerial vehicles (QUAVs) under external disturbances and system uncertainties. First, an asymptotic tracking control strategy is developed to ensure that all tracking errors converge to zero, eliminating steady-state deviations and enhancing operational accuracy. Building on this foundation, a finite-time command-filtered control framework is introduced, which retains the structural advantages of the backstepping approach while guaranteeing rapid convergence within a finite time. Additionally, an error compensation mechanism is incorporated to mitigate filtering errors induced by command filters, further improving control precision. Furthermore, adaptive laws are integrated with fuzzy logic systems to effectively estimate and compensate for the effects from external disturbances and system uncertainties. The effectiveness and feasibility of the proposed method are validated through numerical simulations and experiments on a QUAV prototype.

This paper proposes an adaptive fuzzy output-feedback event-triggered control approach for uncertain discrete-time nonlinear systems with mismatched disturbances. The noncausal problem in traditional discrete-time backstepping can be effectively solved by introducing a command filter into each design step, and the influence of filtering error can be reduced by building an error compensation mechanism. Based on the approximation characteristics of the fuzzy logic systems, the fuzzy state observer is established to obtain the estimation of the immeasurable states although the uncertain dynamics exist. Furthermore, the adaptive update laws are designed for the observer to estimate unknown parameters, and the adaptive controller is developed under the event-triggered mechanism with a dynamic threshold to alleviate the network communication load. The stability analysis of the closed-loop systems is accomplished by constructing weighted Lyapunov functions, excluding Zeno behavior. The validity of the proposed control approach is demonstrated by simulation results.

Accurate correspondence matching in coronary angiography images is crucial for reconstructing 3D coronary artery structures, which is essential for precise diagnosis and treatment planning of coronary artery disease (CAD). Traditional matching methods for natural images often fail to generalize to X-ray images due to inherent differences such as lack of texture, lower contrast, and overlapping structures, compounded by insufficient training data. To address these challenges, we propose a novel pipeline that generates realistic paired coronary angiography images using a diffusion model conditioned on 2D projections of 3D reconstructed meshes from Coronary Computed Tomography Angiography (CCTA), providing high-quality synthetic data for training. Additionally, we employ large-scale image foundation models to guide feature aggregation, enhancing correspondence matching accuracy by focusing on semantically relevant regions and keypoints. Our approach demonstrates superior matching performance on synthetic datasets and effectively generalizes to real-world datasets, offering a practical solution for this task. Furthermore, our work investigates the efficacy of different foundation models in correspondence matching, providing novel insights into leveraging advanced image foundation models for medical imaging applications.

The accurate segmentation of guidewires in interventional cardiac fluoroscopy videos is crucial for computer-aided navigation tasks. Although deep learning methods have demonstrated high accuracy and robustness in wire segmentation, they require substantial annotated datasets for generalizability, underscoring the need for extensive labeled data to enhance model performance. To address this challenge, we propose the Segmentation-guided Frame-consistency Video Diffusion Model (SF-VD) to generate large collections of labeled fluoroscopy videos, augmenting the training data for wire segmentation networks. SF-VD leverages videos with limited annotations by independently modeling scene distribution and motion distribution. It first samples the scene distribution by generating 2D fluoroscopy images with wires positioned according to a specified input mask, and then samples the motion distribution by progressively generating subsequent frames, ensuring frame-to-frame coherence through a frame-consistency strategy. A segmentation-guided mechanism further refines the process by adjusting wire contrast, ensuring a diverse range of visibility in the synthesized image. Evaluation on a fluoroscopy dataset confirms the superior quality of the generated videos and shows significant improvements in guidewire segmentation.

Text data augmentation is an effective strategy for overcoming the challenge of limited sample sizes in many natural language processing (NLP) tasks. This challenge is especially prominent in the few-shot learning (FSL) scenario, where the data in the target domain is generally much scarcer and of lowered quality. A natural and widely used strategy to mitigate such challenges is to perform data augmentation to better capture data invariance and increase the sample size. However, current text data augmentation methods either can’t ensure the correct labeling of the generated data (lacking faithfulness), or can’t ensure sufficient diversity in the generated data (lacking compactness), or both. Inspired by the recent success of large language models (LLM), especially the development of ChatGPT, we propose a text data augmentation approach based on ChatGPT (named ”AugGPT”). AugGPT rephrases each sentence in the training samples into multiple conceptually similar but semantically different samples. The augmented samples can then be used in downstream model training. Experiment results on multiple few-shot learning text classification tasks show the superior performance of the proposed AugGPT approach over state-of-the-art text data augmentation methods in terms of testing accuracy and distribution of the augmented samples.

Recently, ChatGPT and GPT-4 have emerged and gained immense global attention due to their unparalleled performance in language processing. Despite demonstrating impressive capability in various open-domain tasks, their adequacy in highly specific fields like radiology remains untested. Radiology presents unique linguistic phenomena distinct from open-domain data due to its specificity and complexity. Assessing the performance of large language models (LLMs) in such specific domains is crucial not only for a thorough evaluation of their overall performance but also for providing valuable insights into future model design directions: whether model design should be generic or domain-specific. To this end, in this study, we evaluate the performance of ChatGPT/GPT-4 on a radiology natural language inference (NLI) task and compare it to other models fine-tuned specifically on task-related data samples. We also conduct a comprehensive investigation on ChatGPT/GPT-4’s reasoning ability by introducing varying levels of inference difficulty. Our results show that 1) ChatGPT and GPT-4 outperform other LLMs in the radiology NLI task and 2) other specifically fine-tuned Bert-based models require significant amounts of data samples to achieve comparable performance to ChatGPT/GPT-4. These findings not only demonstrate the feasibility and promise of constructing a generic model capable of addressing various tasks across different domains, but also highlight several key factors crucial for developing a unified model, particularly in a medical context, paving the way for future artificial general intelligence (AGI) systems. We release our code and data to the research community.

Vision transformer (ViT) and its variants have achieved remarkable success in various tasks. The key characteristic of these ViT models is to adopt different aggregation strategies of spatial patch information within the artificial neural networks (ANNs). However, there is still a key lack of unified representation of different ViT architectures for systematic understanding and assessment of model representation performance. Moreover, how those well-performing ViT ANNs are similar to real biological neural networks (BNNs) is largely unexplored. To answer these fundamental questions, we, for the first time, propose a unified and biologically plausible relational graph representation of ViT models. Specifically, the proposed relational graph representation consists of two key subgraphs: an aggregation graph and an affine graph. The former considers ViT tokens as nodes and describes their spatial interaction, while the latter regards network channels as nodes and reflects the information communication between channels. Using this unified relational graph representation, we found that: 1) model performance was closely related to graph measures; 2) the proposed relational graph representation of ViT has high similarity with real BNNs; and 3) there was a further improvement in model performance when training with a superior model to constrain the aggregation graph.

Learning with little data is challenging but often inevitable in various application scenarios where the labeled data are limited and costly. Recently, few-shot learning (FSL) gained increasing attention because of its generalizability of prior knowledge to new tasks that contain only a few samples. However, for data-intensive models such as vision transformer (ViT), current fine-tuning-based FSL approaches are inefficient in knowledge generalization and, thus, degenerate the downstream task performances. In this article, we propose a novel mask-guided ViT (MG-ViT) to achieve an effective and efficient FSL on the ViT model. The key idea is to apply a mask on image patches to screen out the task-irrelevant ones and to guide the ViT focusing on task-relevant and discriminative patches during FSL. Particularly, MG-ViT only introduces an additional mask operation and a residual connection, enabling the inheritance of parameters from pretrained ViT without any other cost. To optimally select representative few-shot samples, we also include an active learning-based sample selection method to further improve the generalizability of MG-ViT-based FSL. We evaluate the proposed MG-ViT on classification, object detection, and segmentation tasks using gradient-weighted class activation mapping (Grad-CAM) to generate masks. The experimental results show that the MG-ViT model significantly improves the performance and efficiency compared with general fine-tuning-based ViT and ResNet models, providing novel insights and a concrete approach toward generalizing data-intensive and large-scale deep learning models for FSL.

Medical image segmentation is crucial for clinical decision-making, but the scarcity of annotated data presents significant challenges. Few-shot segmentation (FSS) methods show promise but often require training on the target domain and struggle to generalize across different modalities. Similarly, adapting foundation models such as the segment anything model (SAM) for medical imaging has limitations, including the need for fine-tuning and domain-specific adaptation. To address these issues, we propose a novel method that adapts DINOv2 and SAM 2 for retrieval-augmented few-shot medical image segmentation. Our approach uses DINOv2's feature as query to retrieve similar samples from limited annotated data, which are then encoded as memories and stored in memory bank. With the memory attention mechanism of SAM 2, the model leverages these memories as conditions to generate accurate segmentation of the target image. We evaluated our framework on three medical image segmentation tasks, demonstrating superior performance and generalizability across various modalities without the need for any retraining or fine-tuning. Overall, this method offers a practical and effective solution for few-shot medical image segmentation and holds significant potential as a valuable annotation tool in clinical applications.

In this paper, a discrete-time adaptive fuzzy command filtered backstepping control scheme is presented for the altitude and attitude control problems of the quadrotor unmanned aerial vehicle (UAV). The discrete-time controller is designed by using a novel command filtered backstepping method based on the discrete-time system model of the quadrotor, and model uncertainties are handled by using adaptive fuzzy control. Furthermore, the problem of causality is solved by utilizing the first-order filters. The effectiveness of the proposed control scheme is demonstrated by simulation and experiment.

Preterm birth is the leading cause of death in children under five years old, and is associated with a wide sequence of complications in both short and long term. In view of rapid neurodevelopment during the neonatal period, preterm neonates may exhibit considerable functional alterations compared to term ones. However, the identified functional alterations in previous studies merely achieve moderate classification performance, while more accurate functional characteristics with satisfying discrimination ability for better diagnosis and therapeutic treatment is underexplored. To address this problem, we propose a novel brain structural connectivity (SC) guided Vision Transformer (SCG-ViT) to identify functional connectivity (FC) differences among three neonatal groups: preterm, preterm with early postnatal experience, and term. Particularly, inspired by the neuroscience-derived information, a novel patch token of SC/FC matrix is defined, and the SC matrix is then adopted as an effective mask into the ViT model to screen out input FC patch embeddings with weaker SC, and to focus on stronger ones for better classification and identification of FC differences among the three groups. The experimental results on multi-modal MRI data of 437 neonatal brains from publicly released Developing Human Connectome Project (dHCP) demonstrate that SCG-ViT achieves superior classification ability compared to baseline models, and successfully identifies holistically different FC patterns among the three groups. Moreover, these different FCs are significantly correlated with the differential gene expressions of the three groups. In summary, SCG-ViT provides a powerfully brain-guided pipeline of adopting large-scale and data-intensive deep learning models for medical imaging-based diagnosis.

Visual attention is a fundamental mechanism in the human brain, and it inspires the design of attention mechanisms in deep neural networks. However, most of the visual attention studies adopted eye-tracking data rather than the direct measurement of brain activity to characterize human visual attention. In addition, the adversarial relationship between the attention-related objects and attention-neglected background in the human visual system was not fully exploited. To bridge these gaps, we propose a novel brain-inspired adversarial visual attention network (BI-AVAN) to characterize human visual attention directly from functional brain activity. Our BI-AVAN model imitates the biased competition process between attention-related/neglected objects to identify and locate the visual objects in a movie frame the human brain focuses on in an unsupervised manner. We use independent eye-tracking data as ground truth for validation and experimental results show that our model achieves robust and promising results when inferring meaningful human visual attention and mapping the relationship between brain activities and visual stimuli. Our BI-AVAN model contributes to the emerging field of leveraging the brain's functional architecture to inspire and guide the model design in artificial intelligence (AI), e.g., deep neural networks.

Shortcut learning in deep learning models occurs when unintended features are prioritized, resulting in degenerated feature representations and reduced generalizability and interpretability. However, shortcut learning in the widely used vision transformer (ViT) framework is largely unknown. Meanwhile, introducing domain-specific knowledge is a major approach to rectifying the shortcuts that are predominated by background-related factors. For example, eye-gaze data from radiologists are effective human visual prior knowledge that has the great potential to guide the deep learning models to focus on meaningful foreground regions. However, obtaining eye-gaze data can still sometimes be time-consuming, labor-intensive, and even impractical. In this work, we propose a novel and effective saliency-guided ViT (SGT) model to rectify shortcut learning in ViT with the absence of eye-gaze data. Specifically, a computational visual saliency model (either pretrained or fine-tuned) is adopted to predict saliency maps for input image samples. Then, the saliency maps are used to filter the most informative image patches. Considering that this filter operation may lead to global information loss, we further introduce a residual connection that calculates the self-attention across all the image patches. The experiment results on natural and medical image datasets show that our SGT framework can effectively learn and leverage human prior knowledge without eye-gaze data and achieves much better performance than baselines. Meanwhile, it successfully rectifies the harmful shortcut learning and significantly improves the interpretability of the ViT model, demonstrating the promise of transferring human prior knowledge derived visual saliency in rectifying shortcut learning.

Linking computational natural language processing (NLP) models and neural responses to language in the human brain on the one hand facilitates the effort towards disentangling the neural representations underpinning language perception, on the other hand provides neurolinguistics evidence to evaluate and improve NLP models. Mappings of an NLP model’s representations of and the brain activities evoked by linguistic input are typically deployed to reveal this symbiosis. However, two critical problems limit its advancement: 1) The model’s representations (artificial neurons, ANs) rely on layer-level embeddings and thus lack fine-granularity; 2) The brain activities (biological neurons, BNs) are limited to neural recordings of isolated cortical unit (i.e., voxel/region) and thus lack integrations and interactions among brain functions. To address those problems, in this study, we 1) define ANs with fine-granularity in transformer-based NLP models (BERT in this study) and measure their temporal activations to input text sequences; 2) define BNs as functional brain networks (FBNs) extracted from functional magnetic resonance imaging (fMRI) data to capture functional interactions in the brain; 3) couple ANs and BNs by maximizing the synchronization of their temporal activations. Our experimental results demonstrate 1) The activations of ANs and BNs are significantly synchronized; 2) the ANs carry meaningful linguistic/semantic information and anchor to their BN signatures; 3) the anchored BNs are interpretable in a neurolinguistic context. Overall, our study introduces a novel, general, and effective framework to link transformer-based NLP models and neural activities in response to language and may provide novel insights for future studies such as brain-inspired evaluation and development of NLP models.

Over decades, neuroscience has accumulated a wealth of research results in the text modality that can be used to explore cognitive processes. Meta-analysis is a typical method that successfully establishes a link from text queries to brain activation maps using these research results, but it still relies on an ideal query environment. In practical applications, text queries used for meta-analyses may encounter issues such as semantic redundancy and ambiguity, resulting in an inaccurate mapping to brain images. On the other hand, large language models (LLMs) like ChatGPT have shown great potential in tasks such as context understanding and reasoning, displaying a high degree of consistency with human natural language. Hence, LLMs could improve the connection between text modality and neuroscience, resolving existing challenges of meta-analyses. In this study, we propose a method called Chat2Brain that combines LLMs to basic text-2-image model, known as Text2Brain, to map open-ended semantic queries to brain activation maps in data-scarce and complex query environments. By utilizing the understanding and reasoning capabilities of LLMs, the performance of the mapping model is optimized by transferring text queries to semantic queries. We demonstrate that Chat2Brain can synthesize anatomically plausible neural activation patterns for more complex tasks of text queries.

This article investigates the consensus tracking of uncertain multiple robot manipulators with disturbances, in which a distributed adaptive gain-varying finite-time event-triggered strategy is given. To build the connections between error information and control gains, the well-designed dynamic gain functions are added to the static gains; then, the antidisturbance ability of the networked system is strengthened when there are strong disturbances. The command filters are employed to avoid the direct differentials of virtual controllers, and the compensation strategy with dynamic gains removes the filtering errors compared with traditional backstepping-based algorithms. The event-triggered mechanism is further used to reduce the waste of communication resources by avoiding frequent controller updating for each manipulator in the complex system. Three simulation examples are used to demonstrate its effectiveness.

Learning harmful shortcuts such as spurious correlations and biases prevents deep neural networks from learning meaningful and useful representations, thus jeopardizing the generalizability and interpretability of the learned representation. The situation becomes even more serious in medical image analysis, where the clinical data are limited and scarce while the reliability, generalizability and transparency of the learned model are highly required. To rectify the harmful shortcuts in medical imaging applications, in this paper, we propose a novel eye-gaze-guided vision transformer (EG-ViT) model which infuses the visual attention from radiologists to proactively guide the vision transformer (ViT) model to focus on regions with potential pathology rather than spurious correlations. To do so, the EG-ViT model takes the masked image patches that are within the radiologists' interest as input while has an additional residual connection to the last encoder layer to maintain the interactions of all patches. The experiments on two medical imaging datasets demonstrate that the proposed EG-ViT model can effectively rectify the harmful shortcut learning and improve the interpretability of the model. Meanwhile, infusing the experts' domain knowledge can also improve the large-scale ViT model's performance over all compared baseline methods with limited samples available. In general, EG-ViT takes the advantages of powerful deep neural networks while rectifies the harmful shortcut learning with human expert's prior knowledge. This work also opens new avenues for advancing current artificial intelligence paradigms by infusing human intelligence.

This brief is concerned with neural network (NN)-based adaptive finite-time output feedback attitude tracking control for rigid spacecraft in the presence of actuator saturation, inertial uncertainty, and external disturbance. First, a neural state observer is designed to estimate the unknown state. Then, based on the estimated state, the adaptive neural finite-time command filtered backstepping (CFB) is applied to construct virtual control signal and controller with updating law. The finite-time command filter is given to avoid the computation complexity problem in traditional backstepping, and the compensation signals based on fractional power are constructed to remove filtering errors. Using Lyapunov stability theory, we show that the attitude tracking error (TE) can converge into the desired neighborhood of the origin in finite time and all the signals in the closed-loop system are bounded in finite time although input saturation exists. The numerical simulations are used to show the effectiveness of the given algorithm.

In this paper, the observer based neural adaptive finite-time containment control strategy for non-strict feedback nonlinear multi-agent systems is studied. The finite-time command filter is used to overcome the explosion of complexity problem and the established fractional power based error compensation signal is applied to compensate the filtering error caused by the filter. The distributed finite-time command filtered backstepping control method combines with the neural adaptive control technology and state observer is given, which ensures the containment control errors reach to the desired neighborhood of the origin in finite-time in the presence of uncertain dynamics and unmeasurable states in the system. The given numerical simulations show the effectiveness of the proposed control strategy.

This article is concerned with the containment control of multiple manipulators with uncertain parameters. A novel distributed adaptive backstepping strategy is given in the finite-time control framework. The finite-time command filters (FTCFs) used in the strategy can avoid the explosion of complexity problem for conventional backstepping. To further improve the control performance, the filtering errors caused by the used FTCFs are removed by using the error compensation mechanism (ECM). The proposed virtual control signal, the control torque, and the adaptive updating law can guarantee the set tracking errors converge to an adjustable neighborhood of the origin in finite time in the presence of uncertain parameters. Because the virtual control signal and ECM only use the local information, the established method is completely distributed. Two simulation examples are given to show the effectiveness of the proposed scheme.

In this article, a new command filtered backstepping based finite-time adaptive fuzzy tracking control scheme for a class of unknown nonlinear systems with full-state constraints is established. First, the proposed finite-time command filter will filtering the virtual control signal and get the intermediate control signal within finite-time, so the problem of calculating complexity will not occur in the backstepping process. Then, the fraction-power-based error compensate signal is set up, which can eliminate the influence of filtering error on the control performance. Considering that the unknown nonlinearities exist in the system, the fuzzy logic system based adaptive control technique is used to deal with them, and only one parameter needs to be estimated. It is shown that the states will not violate the prescribed constrains, all the signals in the closed-loop system are bounded in finite-time and the tracking error can converge to the desired neighborhood of the origin in finite time under the barrier Lyapunov function and fraction-power-based virtual control signals. Finally, the effectiveness of the control method is shown by the simulations.

This article is concerned with the tracking control problem for uncertain high-order nonlinear systems in the presence of input saturation. A finite-time control strategy combined with neural state observer and command filtered backstepping is proposed. The neural network models the unknown nonlinear dynamics, the finite-time command filter (FTCF) guarantees the approximation of its output to the derivative of virtual control signal in finite time at the backstepping procedure, and the fraction power-based error compensation system compensates for the filtering errors between FTCF and virtual signal. In addition, the input saturation problem is dealt with by introducing the auxiliary system. Overall, it is shown that the designed controller drives the output tracking error to the desired neighborhood of the origin at a finite time and all the signals in the closed-loop system are bounded at a finite time. Two simulation examples are given to demonstrate the control effectiveness.

This article discusses the distributed continuous-time optimization problem (DCTOP) of second-order multiagent systems (SOMASs). It is assumed that the inputs are required to be in some nonconvex sets, the team objective function (TOF) is a combination of general differentiable convex functions, and each agent can only obtain the information of one local objective function. Based on the neighbors’ information, a new distributed continuous-time optimization algorithm (DCTOA) is first proposed for each agent, where its gradient gains are nonuniform. By introducing a scaling factor and a model transformation, the corresponding system is changed into a time-varying nonlinear system which does not contain constraint operator in form. Then, it is proven that all agents’ states could reach an agreement and the TOF could be minimized by constructing some new Lyapunov functions. Finally, the effectiveness of the algorithm is shown by simulation results.

In this paper, the problem of adaptive finite-time attitude containment control for multiple spacecrafts with unknown external disturbances in spacecraft formation flying is investigated. A distributed control strategy combined with finite-time command filtered backstepping (FTCFB) and an adaptive technique is proposed, which can guarantee the containment errors of attitudes between leader spacecrafts and follower spacecrafts reaching the desired neighborhood in finite time. Moreover, the applied second-order sliding mode differentiator in FTCFB can make the output of the command filter fast approximate the derivative of the virtual signal at the second step of backstepping, which can further improve the control quality. A simulation example is given to show the effectiveness of the new designed technique presented.

This paper addresses a class of network-to-network control problems in the presence of heterogeneous topologies. To achieve the coordination of nodes in a controlled network, a networked controller with a heterogeneous topology is presented in an observer form based on the nearest neighbor rule. It is shown that the network-to-network control problem can be transformed into a coordination problem on a network subject to hybrid static and dynamic interactions. Furthermore, these hybrid interactions among nodes can be represented by appropriately constructing a dynamic graph, of which the connectivity provides a necessary and sufficient guarantee for all nodes to reach agreement on the average quantity of their initial conditions. Simulations are given to illustrate the effectiveness of our results obtained through the dynamic graph approach.

This paper considers the fuzzy finite-time tracking control problem for a class of nonlinear systems with input saturation. A novel fuzzy finite-time command filtered backstepping approach is proposed by introducing the fuzzy finite-time command filter, designing the new virtual control signals and the modified error compensation signals. The proposed approach not only holds the advantages of the conventional command-filtered backstepping control, but also guarantees the finite-time convergence. A practical example is included to show the effectiveness of the proposed method.

This paper is concerned with the finite-time consensus tracking control problems of uncertain nonlinear multiagent systems. A neural network-based distributed adaptive finite-time control scheme is developed, which can guarantee the consensus tracking is achieved in finite time with sufficient accuracy in the presence of unknown mismatched nonlinear dynamics. Such a finite-time feature is achieved by the modified command filtered backstepping technique based on the high-order sliding mode differentiator. Moreover, the proposed control scheme is completely distributed, since the control laws only use the local information. In addition, although mismatched uncertainty nonlinear dynamics are considered, only one parameter needs to be updated for each agent in the control scheme, which will simply the computations and make the proposed scheme more effective for applications. An example is included to verify the presented method.

In this study, an effective robust transcale estimation algorithm is proposed for discrete‐time systems, which are observed by a single sensor at the finest resolution or by two sensors at the finest and coarsest resolutions. The discrete‐time state‐space models of approximation and detail coefficients at each resolution are established by using Haar wavelet decomposition, respectively. The algorithm is developed based on the standard H ∞ filtering scheme, and hence preserves the merits of the H ∞ filter for random signal estimation in the sense that it minimises the effect of the worst possible disturbances on the estimation errors. The proposed algorithm is demonstrated through Monte Carlo simulations involving tracking of a target in CV model.