As cities around the world aim to improve walkability and safety, understanding the irregular and unpredictable nature of pedestrian behavior has become increasingly important. This study introduces a data-driven framework for modeling chaotic pedestrian movement using empirically observed trajectory data and supervised learning. Videos were recorded during both daytime and nighttime conditions to capture pedestrian dynamics under varying ambient and traffic contexts. Pedestrian trajectories were extracted through computer vision techniques, and behavioral chaos was quantified using four chaos metrics: Approximate Entropy and Lyapunov Exponent, each computed for both velocity and direction change. A Principal Component Analysis (PCA) was then applied to consolidate these indicators into a unified chaos score. A comprehensive set of individual, group-level, and contextual traffic features was engineered and used to train Random Forest and CatBoost regression models. CatBoost models consistently achieved superior performance. The best daytime PCA-based CatBoost model reached an R^2 of 0.8319, while the nighttime PCA-based CatBoost model attained an R^2 of 0.8574. SHAP analysis highlighted that features such as distance travel, movement duration, and speed variability were robust contributors to chaotic behavior. The proposed framework enables practitioners to quantify and anticipate behavioral instability in real-world settings. Planners and engineers can use chaos scores to identify high-risk pedestrian zones, apprise infrastructure improvements, and calibrate realistic microsimulation models. The approach also supports adaptive risk assessment in automated vehicle systems by capturing short-term motion unpredictability grounded in observable, interpretable features.

Existing research regarding users' trust in automation has identified associations between trust and users' personality dimensions, including neuroticism and extraversion ( Merritt et al., 20 08; Hoff et al., 20 15) . Additionally, PI, which reflects an individual's tendency to accept novel experienc es and technologies, has been found to be associated with openness to A V adoption ( Deb et al., 2017) . On the other hand, e xisting research on pedestrian - A V interactions primarily relied on laboratory simulations or closed - road field studies, which, though, has provide d valuable theoretical insights (Chang et al., 2024; Clamann et al., 2017; Ma-hadevan et al., 2019), often oversimplif ied the complexity of real - world traffic environments by neglecting the dynamic and unpredictable characteristics of urban ecosystems (Beggiato et al., 2017) and nullified the risks that pedestrians may face in real - world traffic scenarios . As a result, there is still a gap regarding how pedestrians ' trust in A Vs evolves in dynamic, highly interactive real - world scenarios. Addressing thi s gap is essential for accurately capturing the evolution of pedestrians' attitudes toward A V technology and providing actionable insights for optimizing A V design. Thus, a user experiment was conducted at a real - world uncontrolled urban road intersection where participants interacted with commercially running robot taxis that can be categorized as Level 4 by the S ociety of A utomotive E ngineers (SAE) ( SAE international, 2021) . By evaluating participants' perception of A Vs before and after they interacted with the A Vs, for the first time, we explored how interacting with the A Vs can affect the formation of participants' trust in A Vs . Further, given that individual differences, such as personalit y ( Kraus et al., 202 1; Nordhoff et al., 202 5) and p ersonal i nnovativeness ( Hegner et al., 20 19) have been found to influence users' trust in A Vs, we also explored the factors that can moderate pedestrians' trust in A Vs during the interaction process. METHODS Experimental Site As shown in Figure 1, t he experiment was conducted at an uncontrolled intersection with a zebra crossing but without a traffic light on a straight, two - way urban road .

As autonomous robots become more common in public spaces, spontaneous encounters with laypersons are more frequent. For this, robots need to be equipped with communication strategies that enhance momentary transparency and reduce the probability of critical situations. Adapting these robotic strategies requires consideration of robot movements, environmental conditions, and user characteristics and states. While numerous studies have investigated the impact of distraction on pedestrians' movement behavior, limited research has examined this behavior in the presence of autonomous robots. This research addresses the impact of robot type and robot movement pattern on distracted and undistracted pedestrians' movement behavior. In a field setting, unaware pedestrians were videotaped while moving past two working, autonomous cleaning robots. Out of N=498 observed pedestrians, approximately 8% were distracted by smartphones. Distracted and undistracted pedestrians did not exhibit significant differences in their movement behaviors around the robots. Instead, both the larger sweeping robot and the offset rectangular movement pattern significantly increased the number of lateral adaptations compared to the smaller cleaning robot and the circular movement pattern. The offset rectangular movement pattern also led to significantly more close lateral adaptations. Depending on the robot type, the movement patterns led to differences in the distances of lateral adaptations. The study provides initial insights into pedestrian movement behavior around an autonomous cleaning robot in public spaces, contributing to the growing field of HRI research.

Pedestrian intention prediction is essential for autonomous driving in complex urban environments. Conventional approaches depend on supervised learning over frame sequences and require extensive retraining to adapt to new scenarios. Here, we introduce BF-PIP (Beyond Frames Pedestrian Intention Prediction), a zero-shot approach built upon Gemini 2.5 Pro. It infers crossing intentions directly from short, continuous video clips enriched with structured JAAD metadata. In contrast to GPT-4V based methods that operate on discrete frames, BF-PIP processes uninterrupted temporal clips. It also incorporates bounding-box annotations and ego-vehicle speed via specialized multimodal prompts. Without any additional training, BF-PIP achieves 73% prediction accuracy, outperforming a GPT-4V baseline by 18 %. These findings illustrate that combining temporal video inputs with contextual cues enhances spatiotemporal perception and improves intent inference under ambiguous conditions. This approach paves the way for agile, retraining-free perception module in intelligent transportation system.

V2P Collision Warnings for Distracted Pedestrians: A Comparative Study with Traditional Auditory Alerts Novel Certad ID Graduate Student Member, IEEE, Enrico Del Re ID Student Member, IEEE, Joshua V arughese ID Member, IEEE, and Cristina Olaverri-Monreal ID Senior Member, IEEE Abstract -- This study assesses a V ehicle-to-Pedestrian (V2P) collision warning system compared to conventional vehicle-issued auditory alerts in a real-world scenario simulating a vehicle on a fixed track, characterized by limited maneuverability and the need for timely pedestrian response. The results from analyzing speed variations show that V2P warnings are particularly effective for pedestrians distracted by phone use (gaming or listening to music), highlighting the limitations of auditory alerts in noisy environments. The findings suggest that V2P technology offers a promising approach to improving pedestrian safety in urban areas I. I NTRODUCTION Road traffic accidents are a significant global concern, with a disproportionate number of fatalities and injuries affecting Vulnerable Road Users (VRUs) [1]. Among the various factors contributing to these accidents, pedestrian distraction, particularly due to smartphone use, has become a critical issue. Studies have shown that a substantial percentage of pedestrians engage with their smartphones while walking, leading to reduced situational awareness, increased risky behavior, and a higher likelihood of near collisions and accidents [1] [2].

The increasing use of robots in human-centric public spaces such as shopping malls, sidewalks, and hospitals, requires understanding of how pedestrians respond to their presence. However, existing research lacks comprehensive datasets that capture the full range of pedestrian behaviors, e.g., including avoidance, neutrality, and attraction in the presence of robots. Such datasets can be used to effectively learn models capable of accurately predicting diverse responses of pedestrians to robot presence, which are crucial for advancing robot navigation strategies and optimizing pedestrian-aware motion planning. In this paper, we address these challenges by collecting a novel dataset of pedestrian motion in two outdoor locations under three distinct conditions, i.e., no robot presence, a stationary robot, and a moving robot. Thus, unlike existing datasets, ours explicitly encapsulates variations in pedestrian behavior across the different robot conditions. Using our dataset, we propose a novel Neural Social Robot Force Model (NSRFM), an extension of the traditional Social Force Model that integrates neural networks and robot-induced forces to better predict pedestrian behavior in the presence of robots. We validate the NSRFM by comparing its generated trajectories on different real-world datasets. Furthermore, we implemented it in simulation to enable the learning and benchmarking of robot navigation strategies based on their impact on pedestrian movement. Our results demonstrate the model's effectiveness in replicating real-world pedestrian reactions and its its utility in developing, evaluating, and benchmarking social robot navigation algorithms.

Predicting pedestrian crossing behavior is important for intelligent traffic systems to avoid pedestrian-vehicle collisions. Most existing pedestrian crossing behavior models are trained and evaluated on datasets collected from a single country, overlooking differences between countries. To address this gap, we compared pedestrian road-crossing behavior at unsignalized crossings in Germany and Japan. We presented four types of machine learning models to predict gap selection behavior, zebra crossing usage, and their trajectories using simulator data collected from both countries. When comparing the differences between countries, pedestrians from the study conducted in Japan are more cautious, selecting larger gaps compared to those in Germany. We evaluate and analyze model transferability. Our results show that neural networks outperform other machine learning models in predicting gap selection and zebra crossing usage, while random forest models perform best on trajectory prediction tasks, demonstrating strong performance and transferability. We develop a transferable model using an unsupervised clustering method, which improves prediction accuracy for gap selection and trajectory prediction. These findings provide a deeper understanding of pedestrian crossing behaviors in different countries and offer valuable insights into model transferability.

Predicting pedestrian behavior is challenging yet crucial for applications such as autonomous driving and smart city. Recent deep learning models have achieved remarkable performance in making accurate predictions, but they fail to provide explanations of their inner workings. One reason for this problem is the multi-modal inputs. To bridge this gap, we present Sparse Prototype Network (SPN), an explainable method designed to simultaneously predict a pedestrian's future action, trajectory, and pose. SPN leverages an intermediate prototype bottleneck layer to provide sample-based explanations for its predictions. The prototypes are modality-independent, meaning that they can correspond to any modality from the input. Therefore, SPN can extend to arbitrary combinations of modalities. Regularized by mono-semanticity and clustering constraints, the prototypes learn consistent and human-understandable features and achieve state-of-the-art performance on action, trajectory and pose prediction on TITAN and PIE. Finally, we propose a metric named Top-K Mono-semanticity Scale to quantitatively evaluate the explainability. Qualitative results show the positive correlation between sparsity and explainability. Code available at https://github.com/Equinoxxxxx/SPN.

The presence of robots amongst pedestrians affects them causing deviation to their trajectories. Existing methods suffer from the limitation of not being able to objectively measure this deviation in unseen cases. In order to solve this issue, we introduce a simulation framework that repetitively measures and benchmarks the deviation in trajectory of pedestrians due to robots driven by different navigation algorithms. We simulate the deviation behavior of the pedestrians using an enhanced Social Force Model (SFM) with a robot force component that accounts for the influence of robots on pedestrian behavior, resulting in the Social Robot Force Model (SRFM). Parameters for this model are learned using the pedestrian trajectories from the JRDB dataset [1]. Pedestrians are then simulated using the SRFM with and without the robot force component to objectively measure the deviation to their trajectory caused by the robot in 5 different scenarios. Our work in this paper is a proof of concept that shows objectively measuring the pedestrian reaction to robot is possible. We use our simulation to train two different RL policies and evaluate them against traditional navigation models.

The advancement of socially-aware autonomous vehicles hinges on precise modeling of human behavior. Within this broad paradigm, the specific challenge lies in accurately predicting pedestrian's trajectory and intention. Traditional methodologies have leaned heavily on historical trajectory data, frequently overlooking vital contextual cues such as pedestrian-specific traits and environmental factors. Furthermore, there's a notable knowledge gap as trajectory and intention prediction have largely been approached as separate problems, despite their mutual dependence. To bridge this gap, we introduce PTINet (Pedestrian Trajectory and Intention Prediction Network), which jointly learns the trajectory and intention prediction by combining past trajectory observations, local contextual features (individual pedestrian behaviors), and global features (signs, markings etc.). The efficacy of our approach is evaluated on widely used public datasets: JAAD and PIE, where it has demonstrated superior performance over existing state-of-the-art models in trajectory and intention prediction. The results from our experiments and ablation studies robustly validate PTINet's effectiveness in jointly exploring intention and trajectory prediction for pedestrian behaviour modelling. The experimental evaluation indicates the advantage of using global and local contextual features for pedestrian trajectory and intention prediction. The effectiveness of PTINet in predicting pedestrian behavior paves the way for the development of automated systems capable of seamlessly interacting with pedestrians in urban settings.