Addressing navigation safety challenges posed by restricted harbour channels and dense traffic, this paper proposes an adaptive ship navigation decision-making method under multiple constraints, using the southern waters of Hong Kong's Victoria Harbour as a case study. First, a digital traffic environment is constructed to enable real-time situational awareness in confined waters. Second, drawing inspiration from human drivers’ cognitive mechanisms, a motion prediction and control framework based on sequential rolling is established. This framework employs a separated architecture for prediction and simulation models, resolving the conflict between rapid decision simulation and simulation execution through a closed-loop feedback mechanism. Building upon this, a multi-objective optimisation evaluation function is developed that comprehensively considers dynamic collision risks and environmental geometric constraints. Under strict adherence to the 1972 International Regulations for Preventing Collisions at Sea (COLREGs) and sound seamanship principles, it simultaneously solves for optimal collision avoidance strategies and recovery plans. Case studies using real historical Automatic Identification System (AIS) datasets demonstrate that the proposed method dynamically adapts to residual system errors and changes in target ship course/speed, achieving safe avoidance under multiple constraints while promptly tracking the route (resuming navigation) after completing the clear-way maneuver.

AFGLFNet: A novel multi-sensor fusion approach for predictive human lower-limb locomotion intention

A spectral-temporal Transformer framework for very short-term ship motion prediction with industrial applicability and cross-domain transferability

Improving long-term human motion prediction for physical exercise actions using a lightweight approach

Ship motion prediction is essential in marine engineering, but missing data caused by sensor faults or signal interruptions often degrades the accuracy of long short-term memory (LSTM) models. This study investigates how different missing data rates and imputation methods affect LSTM prediction performance. A ship-motion dataset under various speeds and wave conditions was used to examine model feasibility and hyperparameter sensitivity. Traditional filling strategies, including zero and mean filling, were compared under missing data scenarios. Results show that data loss significantly reduces prediction accuracy. The mean-filling method generally performs better than zero-filling, though its effectiveness decreases with higher data diversity. Proper data clustering can effectively enhance its performance.

Despite substantial progress in decoding biosignals for human motion prediction, the influence of participant- and experiment-related factors on the decodability of these signals has received comparatively little attention. This study evaluates the continuous prediction of hand and wrist joint flexion using the MyoKi database, which comprises surface electromyography, inertial measurement units, and force myography data from 35 participants without disabilities performing 74 daily-life tasks. Unlike existing datasets, MyoKi includes tasks that mimic real-world scenarios by allowing natural movement variations and muscle fatigue. Using a long short-term memory neural network, the impact of participant- and experiment-related factors on decoding accuracy was investigated. Our results show that both expanding sensor coverage to additional muscle regions and combining multiple sensor modalities significantly improve decoding performance, with the greatest gains observed for joints controlled by extrinsic muscles. Muscle fatigue, recording time, and participant characteristics such as weight also influenced model accuracy. However, decoding of movements driven by intrinsic hand muscles remains challenging due to anatomical limitations. These findings highlight the importance of sensor placement and multimodal fusion for myoelectric decoding and provide guidance for optimizing sensor configurations in future prosthetic and robotic applications.

Balancing global coherence and hand-level detail: Frequency-decomposed whole-body human motion prediction

Online parameter identification and dynamic model reconstruction for AUV based on adaptive extended Kalman filter and prediction error method.

CardioMorphNet: Cardiac motion prediction using a shape-guided Bayesian recurrent deep network.

HighlightsWhat are the main findings?Dynamic Occlusion State Prediction via Transformer: We propose a Transformer-based framework that explicitly forecasts the future occlusion ratio of targets by modeling historical trends. By integrating this prediction as a dynamic weighting factor into the loss function, our model adaptively learns to mitigate the impact of varying occlusion severities, significantly enhancing tracking stability during continuous state changes.Physical-Constraint Reasoning via GNN: We develop a GNN-based module that leverages road occupancy and neighboring vehicle poses to infer the existence and motion patterns of targets within occluded regions. This module constructs a heterogeneous graph to determine scene physics, effectively linking fragmented trajectories by generating “virtual perception” states for invisible targets.What are the implications of the main findings?Implication of Dynamic Occlusion State Prediction: Introducing predicted occlusion states into the optimization objective shifts the paradigm from passive reaction to proactive anticipation. This implies that explicitly modeling the dynamics of visibility is crucial for robust tracking, as it allows the system to maintain trajectory continuity even when targets undergo rapid and severe transitions between visible and occluded states.Implication of Physical-Constraint Reasoning: Inferring target existence through road occupancy and neighbor interactions demonstrates that scene priors can effectively substitute for missing sensory data. This creates a “reasoning-based tracking” capability that overcomes the physical limitations of roadside sensors, ensuring high-precision association and minimizing ID switches even in fully occluded or “blind” spots.Despite their extended detection ranges and superior precision compared with onboard sensors, roadside perception systems suffer from severe occlusion artifacts in complex traffic, causing significant tracking failures and ID switches. To address this, we propose a novel Dynamic Occlusion–Predictive Neural Network tailored to challenging roadside environments. First, we introduce a Transformer-based Dynamic Occlusion State Predictor to explicitly model the temporal evolution of occlusion. Unlike traditional tracking methods, this module continuously forecasts future occlusion ratios for each target by analyzing historical occlusion patterns. Critically, these predictions are integrated into the tracking framework as dynamic weighting factors in the loss function, enabling the model to adaptively penalize tracking errors based on the predicted occlusion severity and significantly enhancing robustness against dynamic occlusion scenarios. Second, leveraging the predicted occlusion states, we propose a GNN-based Spatial Reasoning Module to address trajectory fragmentation. This module constructs a heterogeneous graph integrating road occupancy information and neighboring vehicle poses to infer the existence and motion patterns of targets within occluded regions. By analyzing scene-level physical constraints, it generates motion predictions for invisible targets and links these inferred states to fragmented trajectories, ensuring temporally continuous tracking even during prolonged visual occlusions. Experiments on the DAIR-V2X and our self-collected roadside dataset show that our framework outperforms state-of-the-art methods in precision and robustness, achieving a 5.1% MOTA gain over the best baseline. This advantage peaks under high occlusion, where preserving ID continuity and minimizing failures validates its efficacy for real-world roadside multi-target tracking.

Prediction of ship motions based on hybrid ACO–PSO and CNN-BiLSTM-SAT algorithm

Research on modeling and prediction of horizontal maneuvering motion of underwater vehicles based on CFD

Numerical prediction of wave loads and motions for ships with forward speed by a frequency-domain harmonic polynomial method

Artificial intelligence for real-time motion tracking in MRI-guided radiotherapy: A systematic review.

Perception is an active process guided by prior knowledge and expectations, allowing the brain to optimize sensory processing and reduce uncertainty. Although modulations in alpha (7-13 Hz) and beta (15-25 Hz) frequencies have been linked to perceptual and top-down processing, their distinct roles within the predictive coding framework remain elusive. Here, we recorded electroencephalographic activity in 60 participants before (at rest) and during a representational momentum task, where prior information of motion typically biases subsequent perceptual judgments. Within participants, instantaneous frequency modulations occurred throughout the trial. Beta-band modulations tracked the inducer's speed and participant's perceptual choices, consistent with a role in encoding predictive information about stimulus dynamics. In contrast, alpha-band frequency modulations were selectively related to perceptual outcomes in trials that elicited representational momentum, with faster instantaneous alpha frequencies associated with the absence of predictive motion extrapolation. This pattern extends previous findings suggesting that alpha frequency dynamics bias the balance between temporal integration and segregation in the context of motion extrapolation, such that faster alpha rhythms favor veridical segmentation of sensory input, whereas slower alpha rhythms promote predictive integration over time. In contrast to ongoing oscillatory activity, resting-state individual alpha and beta frequencies were not associated with the representational momentum phenomenon, indicating that dynamic, task-related frequency modulations, rather than trait-like oscillatory fingerprints, are critical for understanding predictive perceptual biases in highly dynamic contexts.

In complex terrain, quadruped robot trajectories are susceptible to foot-to-ground contact transients and multi-scale temporal dependencies. Traditional Transformers struggle to simultaneously characterize short-term impacts and long-range gait patterns under real-time constraints. To address this, this paper proposes an improved Transformer temporal modeling algorithm for motion prediction in complex terrain. It systematically integrates multi-scale hybrid attention mechanisms for quadruped trajectory prediction, validating the potential advantages of Transformers in modeling contact transients. The method uses a local-sparse global hybrid attention to capture short-term ground-touching impacts and long-range gait dependencies, adapts non-uniform sampling with relative temporal encoding, and highlights key transient features and fuses information from different time domains in parallel through a dynamic gated feedforward network and a multi-scale temporal encoder. Experiments show that the proposed method achieves a short-term FDE of 0.018m, a reduction of 0.017m compared to the traditional Transformer, and reaches 0.095m and 0.130m in gravel and sloping terrains, respectively, while maintaining an inference latency of 12.45ms. The results show that the proposed framework balances real-time performance and high accuracy in complex terrain and can significantly improve the stability and reliability of future trajectory prediction for quadruped robots.

Motion is an important feature of visual scenes and retinal neuronal circuits selectively signal different motion features. It has been shown that the retina can extrapolate the position of a moving object, thereby compensating sensory transmission delays and enabling signal processing in real-time. Amacrine cells, the inhibitory interneurons of the retina, play essential roles in such computations although their precise function remain unclear. Here, we computationally explore the potential effects of two different inhibitory connectivity motifs on the retina's response to moving objects, in a simplified model of the retina: feed-forward and recurrent feed-back inhibition. In this model, both motifs can account for motion anticipation with two different mechanisms. Feed-forward inhibition truncates motion responses and shifts peak responses forward via subtractive inhibition, whereas recurrent feed-back coupling evokes excitatory and inhibitory waves with different phases that interfere and shift the response peak. A key difference between the two mechanisms is how the anticipatory peak shift scales with the speed of a moving object. Motion prediction with feed-forward circuits monotonically decreases with increasing speeds, while recurrent feed-back coupling induces tuning curves that exhibit a preferred speed for which motion prediction is maximal.

Intensive pig farming presents substantial challenges for individual health monitoring due to high stocking densities, complex occlusion scenarios, and the need for continuous real-time surveillance. Existing monitoring approaches rely heavily on manual inspection, which is labor-intensive and prone to delayed detection of abnormal behaviors and disease symptoms. This study proposes an embedded intelligent monitoring system integrating a pan-tilt gimbal platform with an improved multi-object tracking and anomaly detection framework for automated pig health surveillance. The system employs a modified Periodfill_DeepSORT algorithm that incorporates a ReID network with appearance features and motion prediction trajectories to maintain identity consistency under occlusion and re-entry scenarios. For anomaly detection, a lightweight YOLOv8-based network was trained on 772 abnormal samples across three behavioral categories: movement abnormalities, postural abnormalities, and disease-related abnormalities. Experimental results demonstrate that the Periodfill_DeepSORT algorithm achieves a Multiple Object Tracking Accuracy (MOTA) of 95.34%, a Multiple Object Tracking Precision (MOTP) of 94.77%, and an IDF1 score of 96.88%, with only 12 identity switches across 2000 frames involving 12 targets-27 fewer than the standard DeepSORT algorithm. In occlusion scenarios, MOTA improved from 61.1% to 78.3%. The anomaly detection network achieves an overall detection accuracy of 94.5%, representing an 8.8 percentage point improvement over the baseline model, with recognition accuracies of 96.2% for movement abnormalities, 94.1% for postural abnormalities, and 92.8% for disease-related abnormalities. The system operates at 90 frames per second on embedded hardware with a power consumption of 3.2 watts and a startup time of approximately 1 s, with gimbal angle errors maintained within 3°. These results demonstrate the system's effectiveness and practical feasibility for real-time intelligent health monitoring in intensive livestock farming environments.

Internal anatomical motion challenges precise radiation delivery during external beam radiotherapy. Estimating and compensating for anatomical motion are essential for improving planned dose delivery to target volumes while sparing organs-at-risk. This research achieves accurate motion prediction using only planar X-ray imaging from conventional linear accelerators, without surrogate signals or invasive fiducial markers. We propose Deep-Motion-Net: a patient-specific end-to-end graph neural network (GNN) enabling 3D volumetric organ reconstruction from single in-treatment kV planar X-ray images at arbitrary projection angles. A 2D convolutional neural network (CNN) encoder extracts image features, which four feature pooling networks fuse to a 3D template organ mesh. A ResNet-based graph attention network then deforms the feature-encoded mesh. Training uses synthetically generated organ motion instances and corresponding kV images, created by deforming a reference CT volume aligned with the template mesh, generating digitally reconstructed radiographs (DRRs) at required angles, and applying DRR-to-kV style transfer via conditional CycleGAN. Quantitative testing on synthetic respiratory motion scenarios and qualitative assessment on in-treatment images from four liver cancer patients demonstrated overall mean prediction errors of 0.16±0.13 mm, 0.18±0.19 mm, 0.22±0.34 mm, and 0.12±0.11 mm across datasets. Mean peak prediction errors were 1.39 mm, 1.99 mm, 3.29 mm, and 1.16 mm. This approach leverages accessible in-treatment imaging, avoiding expensive MRI systems or invasive markers. To the best of our knowledge, this is the first deep learning framework reconstructing volumetric 3D organ models from single-view images at arbitrary angles throughout an entire in-treatment scan series. Our approach achieves sub-millimetre accuracy when validated on synthetic motion instances and demonstrates clinical feasibility on real-treatment kV images, for which volumetric ground truth is inherently unavailable. The code is available at https://github.com/isurusuranga/DeepMotionNet .

An accurate ship motion model is the core of the ship autonomous system. Aiming at the problem of limited adaptability of traditional parametric models in actual ship maneuvering motion modelling, this paper proposes a non-parametric modelling method based on the fusion of Bidirectional Temporal Convolutional Network and Long Short-Term Memory network (BiTCN-LSTM). The model collaboratively extracts multi-scale spatio-temporal features and bidirectional coupling characteristics of ship motion through bidirectional temporal convolutional network, uses long short-term memory network to model long-term state dependence, and introduces attention mechanism to adaptively strengthen key feature weights. The Bayesian Optimisation (BO) algorithm is used to perform off-line hyperparameter optimizatio. Using the intelligent research and training vessel ‘Xin Hong Zhuan’ as the research object, comparative experiments show that the proposed model demonstrates significant advantages in ship maneuvering motion prediction, with prediction accuracy superior to that of the comparison models. Additionally, it maintains strong predictive performanc even under untrained maneuvering conditions. The research findings provide powerful technical means for real-time motion forecasting and autonomous decision-making of intelligent ships.