Search Results (107)

Realistic modeling of pedestrian flow in dense public spaces is important for urban crowd geosimulation, mobility analysis, and indoor public-space geo-information modeling. Although prior research has emphasized microscopic agent interactions, higher-level spawn dynamics—governing when and where pedestrians appear—remain less explored, despite their fundamental role in shaping crowd density and flow. Existing approaches often decouple spatial and temporal generation, limiting their ability to capture rich spatio-temporal correlations, and they lack controllability for user-specific scenarios such as high-density environments. In this paper, we propose a Guided Joint Spatio-Temporal Diffusion framework for pedestrian spawn simulation. Our objective is to develop and evaluate a controllable joint spatio-temporal generative model that produces each pedestrian spawn event—its inter-arrival time, origin, and destination—consistent with observed spawn dynamics and a user-specified normalized local spawn-intensity condition. The model addresses the upstream initialization of a crowd simulation, rather than complete trajectory prediction or microscopic interaction simulation, and is evaluated through both next-event accuracy and fixed-horizon controllability. The method leverages spatio-temporal diffusion point processes to jointly model spatial and temporal spawn events, capturing dependencies overlooked by classical and neural point-process-based methods. To support controllable pedestrian-flow generation for geosimulation and downstream applications, we integrate a conditional denoising network with classifier-free guidance, enabling user-specified factors such as crowd density to steer generation. Experiments on the Grand Central dataset demonstrate that our method outperforms strong baselines, reducing temporal error (T-RMSE) by 38% and achieving consistent improvements in spatial and spatio-temporal accuracy. These results show the potential of diffusion-based spatio-temporal modeling for controllable urban crowd geosimulation and pedestrian mobility data generation. Full article

Smartphone-based pedestrian dead reckoning (PDR) provides an infrastructure-free solution for two-dimensional (2D) planar localization in GPS-denied environments, but its open-loop nature makes it sensitive to accumulated step-length and heading errors. These errors grow when pedestrian actions and phone carrying modes change, because conventional methods use a fixed step-length model with a constant Weinberg coefficient. This paper proposes a motion-state-aware PDR method with two key designs. First, a joint motion state defined by action type and carrying mode is recognized from smartphone sensor data using a random-forest classifier. Second, the Weinberg coefficient is modeled through two adaptive variants: a state-wise linear model as the main lightweight adaptation mechanism, and a Transformer-enhanced extension that uses historical step-feature sequences to provide additional temporal smoothing for the per-step coefficient K. Both variants keep the predicted coefficient inside the Weinberg equation to preserve the physical structure of step-length estimation, with offline training minimizing the distance error over each calibrated segment. Heading is estimated by fusing gyroscope increments and magnetometer observations to improve continuity under magnetic disturbance. Experiments on routes with frequent motion-state transitions, including a representative indoor corridor with magnetic disturbance and turns, compare a fixed-parameter baseline and two established adaptive step-length baselines against the proposed variants using coefficient-modeling diagnostics and trajectory-level metrics. More challenging deployments such as underground or multi-floor environments are left for future work. Full article

The Winner-Takes-All (WTA) training paradigm inherently prioritises global trajectory accuracy. For long-tail samples, where terminal prediction is inherently difficult, this prioritisation leads to degraded Final Displacement Error (FDE). This asymmetry is reflected in the considerably poorer performance of FDE than Average Displacement Error (ADE) on such samples, yet its underlying mechanisms remain insufficiently diagnosed in the literature. The core contribution of this paper is a two-stage diagnostic framework that reveals two deeper asymmetries and identifies a previously unrecognised phenomenon, which we term hollow diversity. The framework is designed to be independent of specific model architectures and datasets. We first deployed the framework on the JAAD dataset using a custom-implemented baseline and subsequently validated its effectiveness on the ETH/UCY datasets. Full article

The travel problem of visually impaired people is a worldwide issue that needs urgent attention. Although intelligent guide sticks provide obstacle detection and early warning through multi-sensor fusion and embedded algorithms, existing systems generally cannot model the temporal movement patterns of dynamic obstacles, such as pedestrians and vehicles, thereby hindering intention prediction and active obstacle avoidance. Long short-term memory (LSTM), with its gating mechanism, effectively captures long-term dependencies in trajectories and offers a promising solution. This review compares and analyzes LSTM against other mainstream temporal models under the resource constraints of intelligent guide sticks and finds that LSTM demonstrates a favorable combination in temporal modeling capability, lightweight maturity, and edge deployment feasibility. We categorize five lightweight techniques—architecture simplification, low-rank decomposition, structured pruning, quantization, and knowledge distillation—and examine their compression effectiveness, accuracy preservation, and hardware applicability across typical platforms. Furthermore, this review surveys application cases in speech guidance, trajectory prediction-based obstacle avoidance, positioning and navigation, edge computing, and Internet collaboration, exploring the diverse potential of LSTM in intelligent guide stick scenarios. The findings indicate that, after lightweight processing, LSTM models can meet the deployment requirements of resource-constrained edge devices, suggesting their potential feasibility on resource-constrained hardware platforms. However, existing applications still face challenges in balancing real-time performance and accuracy, meeting stringent resource constraints, and the absence of end-to-end validation on real intelligent guide stick prototypes. The reviewed evidence suggests that LSTM-based prediction represents a promising and practically valuable pathway for transitioning intelligent guide sticks from passive response to active prediction. Future research should prioritize real-world deployment validation, domain-specific data collection, and hardware-software co-design to realize its potential fully. Full article

To address the high storage overhead and inadequate spatial geometric representation associated with raw point cloud data in multi-pedestrian trajectory prediction, a compressed point cloud-based and attention-enhanced trajectory prediction method (CPCAE) is proposed in the paper. First, for input raw point cloud, a lossy compression module is designed, which improves the Depoco framework by introducing a multi-feature extraction component and employing a coordinate decomposition strategy to optimize compression quality and spatial representation. For input video frames of pedestrians, spatial features are extracted using a 2D convolutional network, and dynamic interactions among pedestrians are captured by a Transformer-based encoder. Then, both spatial attention and modal attention mechanisms are incorporated to dynamically balance the contributions of two modal features and precisely identify key regions and positions. Experimental results evaluate the proposed framework from the perspectives of point cloud compression and downstream trajectory prediction. The results demonstrate that compressed point cloud representations can support competitive trajectory prediction performance in CPCAE. Full article

Modern fulfillment centers share floor space with human workers, making warehouse multi-robot task allocation a safety-critical problem. We propose SHARP (Safe Heterogeneous Allocation with Risk Prediction), a Transformer-based constrained reinforcement-learning framework with a closed-form deployment-time safety mask. Under a Gaussian pedestrian belief and fixed closest-approach directions, the mask uses Bonferroni-allocated per-pair CVaR scores; a nonnegative mask score implies a conservative trajectory-level chance constraint under the stated assumptions. We also present an idealized primal–dual surrogate analysis, without claiming global convergence for the nonconvex Transformer/PPO implementation. Expanded experiments use ten training seeds per learned method and deterministic final-checkpoint evaluation on twenty independently generated held-out instances. No statistically significant difference between SHARP and Lagrangian-PPO was detected in any of the four scenarios. The held-out analysis further reveals late-training instability and severe over-conservatism in the dense S40_high scenario. These findings position SHARP as an auditable geometric filtering mechanism, while identifying conservatism and training stability as important limitations for deployment. Full article

Pedestrian trajectory prediction in complex occlusion environments remains a critical challenge for autonomous driving systems. Although high-precision prediction models have achieved notable success, they often entail substantial computational overhead and struggle to maintain both accuracy and physical plausibility under real-world occluded conditions. To address these limitations, this paper proposes OPTP-System, a lightweight prediction framework that integrates YOLOv11 with DeepSORT for robust multi-pedestrian tracking in occluded scenes. An extended Kalman filter (EKF)-based motion prediction module is employed to generate trajectory forecasts, while the EKF-derived prior knowledge guides detection re-searching in occluded regions. Furthermore, feedback from trajectory smoothing refines detection confidence, substantially enhancing the model’s capability for continuous tracking and prediction under severe occlusion. Experimental results under challenging occlusion settings (exceeding 50% occlusion) show that the proposed model reduces ADE and FDE by 30.0% and 29.3%, respectively, compared to state-of-the-art methods. These findings demonstrate that OPTP-System achieves superior prediction accuracy while maintaining computational efficiency, offering a practical solution for reliable pedestrian trajectory prediction in complex traffic environments. Full article

Pedestrian safety at urban intersections requires risk-aware mechanisms that extend beyond binary collision detection toward comparative prioritization among multiple agents. This study introduces the Intelligent Pedestrian Model (IPM), a reference-normalized scalar framework that represents pedestrian risk as a function of trajectory, contextual, infrastructural, and behavioral factors, decomposed into Exposure and Severity components. Building on IPM, the Safety-Prioritized Trajectory Model (SPTM) operationalizes the Exposure component using an observation-only, leakage-free kinematic proxy embedded into a cost-aware negative log-likelihood objective. Evaluation on the ETH/UCY benchmark under a strictly inductive protocol shows that moderate prioritization (β ≈ 1.0) improves best-of-K multimodal performance (ALL FDE@K: 0.979 → 0.970 m) while maintaining mean displacement accuracy within seed-level variability. The results indicate that Exposure-based weighting does not act as a global accuracy enhancer but redistributes predictive capacity toward safety-relevant motion regimes. Validation currently covers two ETH/UCY folds under a controlled inductive protocol, while broader cross-fold evaluation remains for future work. Full article

In mixed pedestrian–vehicle traffic environments, overtaking pedestrians by vehicles is a prevalent and complex human–vehicle interaction scenario. However, this maneuver often leads to accidents, resulting in injuries and fatalities, primarily due to inadequate in frastructure, limited pedestrian safety awareness, and suboptimal driver behavior. To mitigate such accidents and develop active vehicle safety systems and autonomous driving algorithms based on human–vehicle interaction data, it is crucial to investigate the overtaking behavior of human drivers. This study examines driver overtaking behavior under various conditions through driving simulator experiments and evaluates how different experimental variables influence driver performance. Using data from 12 skilled drivers, a risk corridor for vehicles overtaking pedestrians is established and a lateral distance prediction model is developed. Based on this established risk corridor, a vehicle overtaking strategy is proposed. Furthermore, to assess the risk level associated with overtaking pedestrians, pedestrians’ subjective risk perceptions are quantified. The simulation results indicate that the maximum lateral error of the vehicle is approximately 0.14 m, the maximum heading error is about 0.06 radians, and the vehicle’s trajectory during pedestrian overtaking remains within the defined risk corridor. These findings are consistent with the operational characteristics of human drivers. Full article

Pedestrian trajectory prediction is vital for crowd analysis and human–-robot interaction. Recent deep models enhance accuracy by modeling social interactions and scene context, but they often remain opaque and rarely address privacy risks associated with learning individualized motion patterns. We propose CAMP, a Context-Aware, Multimodal, and Privacy-preserving pedestrian trajectory prediction framework designed around a role-aligned multimodal architecture, in which trajectory representations, dynamic scene cues, and explicit spatial interaction constraints are modeled through complementary branches. In CAMP, the trajectory encoder separates shared motion regularities from individualized motion tendencies, the optical-flow encoder captures motion-centric transient scene dynamics, and the potential-field encoder provides an interpretable spatial cost prior for obstacle avoidance and social interaction modeling. A Transformer-based decoder fuses these modalities to predict future trajectory distributions. To reduce the exposure of personalized motion patterns, we apply targeted DP-SGD only to the individual branch during the private fine-tuning stage, while treating the remaining frozen components as post-processing under the stated threat model. Experiments on the ETH/UCY benchmark show that CAMP achieves competitive ADE/FDE performance under the reported setting, while its private variant DP-CAMP maintains a reasonable utility–privacy trade-off across several reported privacy budgets. Full article

Pedestrian trajectory prediction is a fundamental task in autonomous driving and mobile robotics, where accurate forecasting requires modeling of both social interactions and scene-related constraints. However, existing methods typically rely on a fixed interaction modeling strategy, which may be insufficient under heterogeneous crowd densities. To address this limitation, we propose DAGH-Net, a density-adaptive gated hybrid network for pedestrian trajectory prediction. Built upon an SR-LSTM (State Refinement for LSTM) backbone, the proposed framework integrates two complementary reasoning pathways: a data-driven social interaction branch and a hybrid knowledge graph branch that encodes structured relational priors among pedestrians, obstacles, and walkable regions. A local-density-conditioned gating mechanism is further introduced to adaptively fuse these features according to the surrounding crowd condition of each pedestrian. This design helps suppress redundant interaction cues in sparse settings while strengthening socially compliant and scene-consistent reasoning in dense or conflict-prone environments. Experimental results on the ETH (Eidgenössische Technische Hochschule Zürich) and UCY (University of Cyprus) benchmarks, evaluated using Mean Average Displacement (MAD) and Final Average Displacement (FAD), show that DAGH-Net improves the average MAD and FAD by 1.6% and 4.2%, respectively, compared with SR-LSTM. Ablation studies further support the complementary contributions of the hybrid knowledge graph and the density-adaptive gating mechanism. We also discuss the limitations of the current density formulation and benchmark scale, which suggest several directions for future improvement. Full article

With the widespread application of autonomous vehicles (AVs), their dynamic interactions with other road users pose significant challenges to trajectory planning. Previous research on trajectory planning in shared spaces has mainly focused on generating smooth trajectories, while research considering the risks of human–vehicle interactions remains insufficient. Therefore, a risk-considered trajectory planning framework for autonomous vehicles is proposed. This framework includes two modules: pedestrian trajectory prediction and vehicle planning. In the prediction module, Social-STGCNN is used to predict pedestrian trajectories, obtaining a series of trajectories and probabilities, which serve as input to the planning module. To ensure the rationality of trajectory planning, a planning model is established in Frenet coordinates based on a quintic polynomial. Combining Bayesian and equality principles, a risk-considered cost function is designed. Under this framework, the risk value is calculated using the pedestrian trajectory prediction probability, and further Bayesian and equality costs are calculated. Based on the constraints, the trajectory with the minimum cost is solved. To evaluate the rationality of this framework, we designed simulation experiments for five typical high-conflict scenarios: overtaking in the same direction, head-on collision, pedestrian crossing, encountering pedestrians from multiple directions, and turning while encountering pedestrians crossing. Simultaneously, the framework is validated in a real-world environment. The results show that the proposed method can accurately capture pedestrians’ crossing intentions and effectively avoid pedestrians. The trajectory generated in the real environment is highly consistent with that of a driver, and it exhibits excellent adaptability and robustness in high-density mixed traffic environments. Full article

Autonomous cleaning robots operating in semi-structured dynamic environments must execute task-oriented motions while safely interacting with surrounding agents. These agents include pedestrians, vehicles, and other robots. In such environments (e.g., interaction-rich campus environments), reliable self-decision-making requires anticipating the future motions of surrounding agents rather than relying solely on reactive obstacle avoidance. This paper presents a trajectory prediction-enabled self-decision-making framework for autonomous cleaning robots in campus environments. A learning-based multi-agent trajectory prediction model is trained offline using public benchmarks and real-world operational data to capture typical interaction patterns in corridor-following, edge-cleaning, and intersection scenarios. The predicted trajectories are then incorporated as forward-looking priors into the robot’s online decision-making and planning process, enabling prediction-aware yielding, detouring, and task continuation decisions. The proposed framework is evaluated using real-world data-driven scenario reconstruction on a high-fidelity simulation platform that incorporates realistic vehicle dynamics and heterogeneous traffic participants. This evaluation focuses on short-horizon prediction performance and its impact on downstream decision-making stability. The results show that integrating trajectory prediction into the decision-making loop leads to more stable motion behavior and fewer abrupt adjustments in interaction scenarios. Under short-term prediction horizons, the evaluation results show that the proposed model achieves ADERate and FDERate exceeding 90% under predefined error thresholds, while lane-change prediction accuracy remains around 79%. In addition, the robot maintains stable speed tracking with only minor fluctuations under medium-density traffic conditions. Full article

As a typical biomimetic robotic system, quadruped robots replicate the flexible locomotion of quadruped mammals, outperforming wheeled robots in human-centered daily scenarios. To improve the social navigation adaptability of biomimetic quadruped robots in human–robot shared environments, this paper proposes a collision-aware orthogonal steering social force model (COSFM), an enhanced social force model that integrates collision prediction and social norms, inspired by human-like collision avoidance behaviors and social interaction rules. The model addresses key limitations of conventional social force models: delayed responses to dynamic pedestrians and inadequate consideration of pedestrians’ comfort zones. It introduces a time-to-collision prediction mechanism to mimic human predictive decision-making in dynamic social interactions, enhancing the robot’s anticipation of pedestrian motion intentions, and designs an orthogonal steering-based avoidance strategy for four typical human–robot interaction scenarios (head-on encounters, intersecting paths, active overtaking, passive yielding). This strategy replicates humans’ natural priority of lateral steering over abrupt deceleration or retreat, generating socially compliant trajectories aligned with human behavioral expectations. The proposed method is validated via simulation and real-world experiments on a Unitree Aliengo quadruped robot. Results show that the COSFM algorithm achieves a higher navigation success rate and better performance in path length, navigation time, and minimum human-robot distance than existing approaches, while its human-like lateral avoidance priority effectively preserves pedestrians’ psychological comfort zones, demonstrating robust social adaptability and great application potential for biomimetic legged robots. Full article

Real-time, precise monitoring of the number and distribution of indoor personnel is crucial for building safety management, operational optimization, and personnel scheduling. However, narrow entrances and high-density passageways often lead to missed detections, false positives, and tracking failures in pedestrian detection, thereby reducing cross-line counting accuracy. Additionally, edge devices deployed in practical scenarios frequently process multiple video streams simultaneously, resulting in computational resource constraints. To address these challenges, this paper proposes a lightweight, enhanced multi-object pedestrian tracking and counting method tailored for indoor scenarios by optimizing deep learning models. Firstly, modular optimizations are applied to the YOLOv8n model to construct a more lightweight detector, RL_YOLOv8, reducing computational overhead while maintaining accuracy. Secondly, correlated pedestrian auxiliary prediction and pedestrian position change constraints are employed to mitigate ID switching, tracking interruptions, and trajectory jumps in dense scenes. Finally, a buffer zone auxiliary counting strategy is designed to further reduce missed detections of pedestrians crossing lines. Experimental results demonstrate that compared to the original detection-and-tracking-based line-crossing counting method, the improved approach effectively enhances counting accuracy and real-time performance, better meeting the requirements of practical intelligent security and crowd monitoring systems. Full article