Virtual Pedestrians Research Articles

Exploring the challenges of avoiding collisions with virtual pedestrians using a dual-task paradigm in individuals with chronic moderate to severe traumatic brain injury.

Individuals with a moderate-to-severe traumatic brain injury (m/sTBI), despite experiencing good locomotor recovery six months post-injury, face challenges in adapting their locomotion to the environment. They also present with altered cognitive functions, which may impact dual-task walking abilities. Whether they present collision avoidance strategies with moving pedestrians that are altered under dual-task conditions, however, remains unclear. This study aimed to compare between individuals with m/sTBI and age-matched control individuals: (1), the locomotor and cognitive costs associated with the concurrent performance of circumventing approaching virtual pedestrians (VRPs) while attending to an auditory-based cognitive task and; (2) gaze behaviour associated with the VRP circumvention task in single and dual-task conditions. Twelve individuals with m/sTBI (age = 43.3 ± 9.5 yrs; >6 mo. post injury) and 12 healthy controls (CTLs) (age = 41.8 ± 8.3 yrs) were assessed while walking in a virtual subway station viewed in a head-mounted display. They performed a collision avoidance task with VRPs, as well as auditory-based cognitive tasks (pitch discrimination and auditory Stroop), both under single and dual-task conditions. Dual-task cost (DTC) for onset distance of trajectory deviation, minimum distance from the VRP, maximum lateral deviation, walking speed, gaze fixations and cognitive task accuracy were contrasted between groups using generalized estimating equations. In contrast to CTLs who showed locomotor DTCs only, individuals with m/sTBI displayed both locomotor and cognitive DTCs. While both groups walked slower under dual-task conditions, only individuals with m/sTBI failed to modify their onset distance of trajectory deviation and maintained smaller minimum distances and smaller maximum lateral deviation compared to single-task walking. Both groups showed shorter gaze fixations on the approaching VRP under dual-task conditions, but this reduction was less pronounced in the individuals with m/sTBI. A reduction in cognitive task accuracy under dual-task conditions was found in the m/sTBI group only. Individuals with m/sTBI present altered locomotor and gaze behaviours, as well as altered cognitive performances, when executing a collision avoidance task involving moving pedestrians in dual-task conditions. Potential mechanisms explaining those alterations are discussed. Present findings highlight the compromised complex walking abilities in individuals with m/sTBI who otherwise present a good locomotor recovery.

Collision avoidance behaviours while young adults avoid a virtual pedestrian approaching on a 45° angle under attentionally demanding conditions

Individuals rely on visual information to determine when to adapt their behaviours (i.e., by changing path and/or speed) to avoid an approaching object or person. After initiating an avoidance behaviour, individuals may control the space (i.e., minimum clearance distance) between themselves and another person or object. The current study aimed to determine the action strategies of young adults while avoiding a virtual pedestrian approaching along a 45° angle in an attentionally demanding task. Twenty-one young adults (22.9 ± 1.9 yrs., 11 males) were immersed in a virtual environment and were instructed to walk along a 7.5 m path towards a goal located along the midline. Two virtual pedestrians (VP) positioned 2.83 m to the left and right of the midline approached participants on a 45° angle. To manipulate the point at which the participants and the VP would intersect during different trials, the VP approached at one of three speeds: 0.8×, 1.0×, or 1.2× each participants' average walking speed. Participants were instructed to walk to a goal without colliding with the VP while performing the attention task; reporting whether a shape changed above the VPs' heads. Results revealed that young adults did not modulate their timing of avoidance to the approach characteristics of the VP, as they consistently avoided the collision 1.67 s after the VP began moving. However, young adults seem to control how they avoid an oncoming collision by maintaining a consistent safety margin after an avoidance behaviour was initiated.

Brain mechanisms involved in the perception of emotional gait: A combined magnetoencephalography and virtual reality study.

Brain processes associated with emotion perception from biological motion have been largely investigated using point-light displays that are devoid of pictorial information and not representative of everyday life. In this study, we investigated the brain signals evoked when perceiving emotions arising from body movements of virtual pedestrians walking in a community environment. Magnetoencephalography was used to record brain activation in 21 healthy young adults discriminating the emotional gaits (neutral, angry, happy) of virtual male/female pedestrians. Event-related responses in the posterior superior temporal sulcus (pSTS), fusiform body area (FBA), extrastriate body area (EBA), amygdala (AMG), and lateral occipital cortex (Occ) were examined. Brain signals were characterized by an early positive peak (P1;∼200ms) and a late positive potential component (LPP) comprising of an early (400-600ms), middle (600-1000ms) and late phase (1000-1500ms). Generalized estimating equations revealed that P1 amplitude was unaffected by emotion and gender of pedestrians. LPP amplitude showed a significant emotion X phase interaction in all regions of interest, revealing i) an emotion-dependent modulation starting in pSTS and Occ, followed by AMG, FBA and EBA, and ii) generally enhanced responses for angry vs. other gait stimuli in the middle LPP phase. LPP also showed a gender X phase interaction in pSTS and Occ, as gender affected the time course of the response to emotional gait. Present findings show that brain activation within areas associated with biological motion, form, and emotion processing is modulated by emotional gait stimuli rendered by virtual simulations representative of everyday life.

Navigating Virtual Collisions: Insights Into Perception–Action Differences in Children and Young Adults Using a Head-On Avoidance Task

Children tend to make more last-minute locomotor adjustments than adults when avoiding stationary obstacles. The purpose of this study was to compare avoidance behaviors of middle-aged children (10–12 years old) with young adults during a head-on collision course with an approaching virtual pedestrian. Participants were immersed in a virtual environment and completed a perceptual decision-making task in which they had to respond to the future direction of an approaching virtual pedestrian once they disappeared. Following the perceptual task, participants walked along an 8-m pathway toward a goal, while avoiding a collision with a virtual pedestrian who approached along the midline than veered toward the left, right, or continued walking straight. Results revealed that children were able to correctly predict the future directions of the virtual pedestrian similar to adults, albeit at a slower response time (0.55 s vs. 0.33 s). During the action task, children initiated a deviation in their travel path later (i.e., closer to the virtual pedestrian) compared to adults (1.65 s vs. 1.52 s). Children were also more variable in their onset of deviation and time-to-contact. Although children appear to have developed adult-like perceptual abilities and can avoid an approaching virtual pedestrian, children employ riskier avoidance strategies and are highly variable, suggesting middle-aged children are still fine-tuning their perception-action system.

Influence of surgical masks on the avoidance of virtual pedestrians

To combat the COVID-19 pandemic, governments around the world have proposed a series of mitigation strategies. While responses varied across different governing bodies, recommendations such as social distancing and the use of surgical masks were nearly universal. These recommendations, as well as the social anxiety that emerged during the pandemic, are likely to have influenced pedestrian interactions. In this study, we have examined the effect of surgical masks on locomotor circumvention strategies in response to virtual pedestrians. We further explored the relationship between measures of obstacle clearance and feelings of anxiety related to community ambulation in the context of the pandemic. Using virtual reality, locomotor circumvention strategies in response to approaching pedestrians with and without surgical masks were measured in a sample of 11 healthy young individuals. Additionally, a questionnaire was developed and used to gain insights into participants’ behaviours during and after a strict period of restrictions that were in effect before the summer of 2020. Results showed that participants maintained a larger clearance when virtual pedestrians wore a surgical mask. Furthermore, clearance was positively associated with anxiety toward community ambulation in the context of the pandemic. Our findings provide evidence that mask-wearing elicits an increase in physical distancing during pedestrian interactions. Furthermore, results indicate that social context and mental health status influence locomotor outcomes measured in the context of a pedestrian interaction task and highlight the potential of virtual reality simulations to study locomotion in the community setting.

The S-BAN: Insights into the Perception of Shape-Changing Haptic Interfaces via Virtual Pedestrian Navigation

Screen-based pedestrian navigation assistance can be distracting or inaccessible to users. Shape-changing haptic interfaces can overcome these concerns. The S-BAN is a new handheld haptic interface that utilizes a parallel kinematic structure to deliver 2-DOF spatial information over a continuous workspace, with a form factor suited to integration with other travel aids. The ability to pivot, extend and retract its body opens possibilities and questions around spatial data representation. We present a static study to understand user perception of absolute pose and relative motion for two spatial mappings, showing the highest sensitivity to relative motions in the cardinal directions. We then present an embodied navigation experiment in virtual reality (VR). User motion efficiency when guided by the S-BAN was statistically equivalent to using a vision-based tool (a smartphone proxy). Although haptic trials were slower than visual trials, participants’ heads were more elevated with the S-BAN, allowing greater visual focus on the environment.

Effects of age on the visuo-locomotor control used to circumvent a virtual pedestrian with different limb movements

It is known that young adults (YA) circumvent pedestrians differently than inanimate obstacles and that limb movements of the pedestrian influence minimum clearance for predictable pedestrian paths. Although older adults (OA) use more cautious strategies for general pedestrian avoidance compared to YA, how pedestrian movements influence circumvention by OAs is unknown. The aim of this study was to understand how limb movements of a pedestrian with an initially unpredictable trajectory affect circumvention control in younger vs older healthy adults. Fourteen YA and 14 OA (> 70 years) were immersed in a virtual shopping mall and instructed to circumvent a virtual pedestrian (VP) approaching with either normal locomotor movements, upper limbs fixed, lower limbs fixed, or both upper and lower limbs fixed. Onset distance for trajectory deviation, minimum clearance, walking speed, body segment yaw angles and gaze behaviour were analysed. When the VP lacked local limb movements, both age groups initiated their trajectory deviations farther away, but significantly more so for OA. Minimal clearance was unchanged across conditions and similar for both age groups. OA walked slower, produced smaller head and trunk yaw, and visually focused on the VP for a greater percentage of time. Thus, lack of limb movements of another pedestrian resulted in more cautious circumvention control and OA needed more time to process visual information with greater visual attention focused on the VP. Age-related changes could translate to a greater risk of falls in OA populations with reduced balance and mobility that could limit community ambulation.

Effects of Small Talk With a Crowd of Virtual Humans on Users' Emotional and Behavioral Responses.

In this contribution, we empirically investigated the effect of small talk on the users' non-verbal behaviors and emotions when users interacted with a crowd of virtual humans (VHs) with positive behavioral dispositions. Users were tasked with collecting items in a virtual marketplace via natural speech-based dialogue with a crowd of virtual pedestrians and vendors. The users were able to engage in natural speech-based conversation in a predefined corpus of small talk content that covered various commonplace small talk topics such as conversations about the weather, general concerns, and entertainment based on similar real-life situations. For instance, the VHs with the small talk ability would ask the users some simple questions to make small talk or remind the users of their belongings. We conducted a between-subjects empirical evaluation to investigate whether the user behaviors and emotions were different between a small talk condition and a non-small talk condition, and examined gender effects of the participants. We collected objective and subjective measures of the users to analyze users' emotions and social interaction behaviors, when in conversation with VHs that either possessed small-talk capability or not, besides task or goal oriented dialogue capabilities. Our result revealed that the VHs with small talk capability could alter the emotions and non-verbal behaviors of the users. Furthermore, the non-verbal behaviors between female and male participants differed greatly in the presence or absence of small talk.

Impact of dual tasking on gaze behaviour and locomotor strategies adopted while circumventing virtual pedestrians during a collision avoidance task.

We investigated gaze behaviour and collision avoidance strategies in 16 healthy young individuals walking towards a goal while exposed to virtual pedestrians (VRPs) approaching from different directions (left, middle, right). This locomotor task and an auditory-based cognitive task were performed under single and dual-task conditions. Longer gaze fixation durations were observed on the approaching vs. other VRPs, with longer fixations devoted to the upper trunk and head compared to other body segments. Compared to other pedestrian approaches, the middle pedestrian received longer fixations and elicited faster walking speeds, larger onset distances of trajectory devitation and smaller obstacle clearances. Gaze and locomotor behaviours were similar between single and dual-task conditions but dual-task costs were observed for the cognitive task. The longer gaze fixations on approaching vs. other pedestrians suggest that enhanced visual attention is devoted to pedestrians posing a greater risk of collision. Likewise, longer gaze fixations for the middle pedestrians may be due to the greater collision risk entailed by this condition, and/or to the fact that this pedestrian was positioned in front of the end goal. Longer fixations on approaching VRPs' trunk and head may serve the purpose of anticipating their walking trajectory. Finally, the dual-task effects that were limited to the cognitive task suggest that healthy young adults prioritize the locomotor task and associated acquisition of visual information. The healthy patterns of visuomotor behaviour characterized in this study will serve as a basis for comparison to further understand defective collision avoidance strategies in patient populations.

Coordinating Clearance and Postural Reorientation When Avoiding Physical and Virtual Pedestrians.

Community ambulation requires efficient locomotor adaptations to avoid collisions with other pedestrians. Virtual reality (VR) offers the possibility to experimentally manipulate the environment, allowing researchers to safely assess locomotor responses during pedestrian interactions. In this paper, locomotor adjustments in response to interferers approaching from different directions were examined in both a physical and a virtual environment. Further analysis compared the extent to which locomotor adjustments were influenced by different circumvention strategies. To achieve these goals, twelve healthy young participants were assessed while walking towards a target and avoiding pedestrians approaching from the left, middle, or right, in both VR and the real world. Results showed that in VR, participants walked with slower velocities, maintained larger minimum distances, and reached larger trajectory deviations. Additionally, trajectory deviations and postural reorientations were executed earlier in VR. There were additional differences according to whether participants had to give way to a head-on approaching interferer or choose to pass in front or behind a diagonally approaching interferer. Compared to the other circumvention strategies, passing in front was associated with faster walking speeds as well as smaller and later trajectory deviations and postural reorientations. Lastly, while a cephalocaudal sequence of segment reorientation was observed in both environments, no difference between the onset of head and thorax reorientation was observed in VR. In conclusion, obstacle clearance and postural reorientation are modulated by the environment and circumvention strategies. These modulations should be considered when designing experiments or clinical interventions.

Mixed reality LVC simulation: A new approach to study pedestrian behaviour

Despite controlled experiments and computer-based simulations are important tools to studying pedestrian behaviour, the pedestrian data collected by both have limitations. These limitations are manifested in the disconnect between the behavioural characteristics of real-world pedestrians and the way they are represented in these simulated environments. Recent developments in mixed reality and live, virtual, and constructive (LVC) simulation have the potential to make up these limitations. This paper reports on a study and project prototypes for building a client-server system using simulation modelling software tools and mixed reality headsets. We set up a mixed reality space containing real pedestrians and holographic-projection virtual pedestrians in a real building environment. The virtual pedestrians move according to a pedestrian dynamics model and interact with real pedestrians and environments in real-time. Taking evacuation in high risk as an example, we demonstrate how interactions between pedestrians, even pedestrians and environments can be achieved in the space while collecting data on the entire process. This study introduces a new approach to study pedestrian behaviour in buildings: mixed reality LVC simulation.

Monte carlo method for estimating whole-body injury metrics from pedestrian impact simulation results

The goal of the current study was to develop a method to estimate whole-body injury metrics (WBIMs), which measure the overall impact of injuries, using stochastic injury prediction results from a computational human surrogate. First, hospitalized pedestrian data was queried to identify injuries sustained by pedestrians and their frequencies. Second, with consideration for an understanding of injury mechanisms and the capability of the computational human surrogate, the whole-body was divided into 17 body regions. Then, an injury pattern database was constructed for each body region for various maximum abbreviated injury scale (MAIS) levels. Third, a two-step Monte Carlo sampling process was employed to generate N virtual pedestrians with an assigned list of injuries in AIS codes. Then, the expected values of WBIMs such as injury severity score (ISS), probability of death, whole-body functional capacity index (WBFCI), and lost years of life (LYL), were estimated. Lastly, the proposed method was verified using injury information from the inpatient pedestrian database. Also, the proposed method was applied to pedestrian impact simulations with various impact speeds to estimate the probability of death with respect to the impact speed. The probability of death from the proposed method was compared with those from epidemiological studies.The proposed method accurately estimated WBIMs such as ISS and WBFCI using either for a given distribution of injury risk or MAIS levels. The predicted probability of death with respect to the impact speed showed a good correlation with those from the epidemiological study. These results imply that if we have a human surrogate that can predict the risk of injury accurately, we can accurately estimate WBIMs using the proposed method. The proposed method can simplify a vehicle design optimization process by transforming the multi-objective optimization problem into the single-objective one. Lastly, the proposed method can be applied to other human surrogates such as occupant models.

Virtual to Real Adaptation of Pedestrian Detectors.

Pedestrian detection through Computer Vision is a building block for a multitude of applications. Recently, there has been an increasing interest in convolutional neural network-based architectures to execute such a task. One of these supervised networks’ critical goals is to generalize the knowledge learned during the training phase to new scenarios with different characteristics. A suitably labeled dataset is essential to achieve this purpose. The main problem is that manually annotating a dataset usually requires a lot of human effort, and it is costly. To this end, we introduce ViPeD (Virtual Pedestrian Dataset), a new synthetically generated set of images collected with the highly photo-realistic graphical engine of the video game GTA V (Grand Theft Auto V), where annotations are automatically acquired. However, when training solely on the synthetic dataset, the model experiences a Synthetic2Real domain shift leading to a performance drop when applied to real-world images. To mitigate this gap, we propose two different domain adaptation techniques suitable for the pedestrian detection task, but possibly applicable to general object detection. Experiments show that the network trained with ViPeD can generalize over unseen real-world scenarios better than the detector trained over real-world data, exploiting the variety of our synthetic dataset. Furthermore, we demonstrate that with our domain adaptation techniques, we can reduce the Synthetic2Real domain shift, making the two domains closer and obtaining a performance improvement when testing the network over the real-world images.

Contextual sensory integration training via head mounted display for individuals with vestibular disorders: a feasibility study

Purpose Virtual reality (VR) interventions can simulate real-world sensory environments. The purpose of this study was to test the feasibility of a novel VR application (app) developed for a Head Mounted Display (HMD) to target dizziness, imbalance and sensory integration in a functional context for patients with vestibular disorders. Here we describe the design of the app as well as self-reported and functional outcomes in vestibular patients before and after participating in vestibular rehabilitation using the app. Material and methods Our app includes a virtual street, airport, subway or a park. The clinician controls the visual and auditory load including several levels of direction, amount and speed of virtual pedestrians. Clinicians enrolled 28 patients with central (mild-traumatic brain injury [mTBI] or vestibular migraine) and peripheral vestibular disorders. We recorded the Simulator Sickness Questionnaire, Visual Vertigo Analogue Scale (VVAS), Dizziness Handicap Inventory (DHI), Activities-Specific Balance Confidence Scale (ABC), 8-foot up and go (8FUG) and Four-Step Square Test (FSST) before and after the intervention. Results Within the 15 patients who completed the study, 12 with peripheral hypofunction showed significant improvements on the VVAS (p = 0.02), DHI (p = 0.008) and ABC (p = 0.02) and a small significant improvement on the FSST (p = 0.015). Within-session changes in symptoms were minimal. Two patients with mTBI showed important improvements, but one patient with vestibular migraine, did not. Conclusion HMD training within increasingly complex immersive environments appears to be a promising adjunct modality for vestibular rehabilitation. Future controlled studies are needed to establish effectiveness. IMPLICATIONS FOR REHABILITATION Virtual Reality allows for gradual introduction of complex semi-real visual environments. Within VR training patients can re-learn to maintain balance when presented with a sensory conflict in a safe environment. Head Mounted Display training appears to be a promising adjunct modality for vestibular rehabilitation. Portability and affordability of the hardware and software enhance the potential clinical outreach.

Reading text messages at different stages of pedestrian circumvention affects strategies for collision avoidance in young and older adults

BackgroundReading text messages is associated with accidents while walking in community places. Research questionTo what extent does reading text messages at different stages of obstacle circumvention affect avoidance strategies while walking in young vs. older adults? MethodsSixteen healthy young and 14 older adults were assessed while walking and viewing a virtual environment (VE) simulating a subway station with three virtual pedestrians positioned 7.5 m away from the participant in the centre (0°), left and right (±40°). As participants advanced 0.5 m towards a target in the far space, a virtual pedestrian randomly approached them. Text messages were delivered at onset of pedestrian movement (0.5 m; early message) or during obstacle circumvention (2.5 m; late message). A Vicon motion capture system captured trajectory displacement while walking. ResultsIn both age groups, accuracy of message report (AMR) was reduced for early compared to late messages (p < 0.001), although older adults showed larger deterioration (p < 0.001) compared to younger participants. Locomotor outcomes (obstacle clearance, onset time of avoidance and walking speed) showed no differences between young and older participants (p > 0.05). Early messages led to slower walking speed (p < 0.001) and more frequent collisions compared to late messages and the no-message condition. Late messages yielded faster walking speed (p < 0.001) and onset time of avoidance (p < 0.02) compared to the other conditions. SignificanceResults indicate that the stage of an avoidance strategy at which text messages are received impacts on pedestrian circumvention, with early messages posing a greater challenge to collision avoidance. In older adults, the reduced AMR suggests larger dual-task interference and prioritization of the walking task. The lack of further walking speed reduction in older vs. young adults may put them at greater risk of collisions in crowded, unpredictable community environments.

Phone messages affect the detection of approaching pedestrians in healthy young and older adults immersed in a virtual community environment.

BackgroundMobile phones are increasingly associated with accidents while walking. Little is known, however, about the impact of phone messaging on the actual perception of other pedestrians. This study aimed to investigate the extent to which the detection of approaching pedestrians is affected by the sensory modality (text or audio) of phone messages in young vs. older adults.MethodsEighteen healthy young (24 ± 2.9 years) and 15 older adults (68 ± 4.2 years) performed a phone message deciphering task, an obstacle detection task, and a dual-task condition combining both tasks. Participants were tested while seated and viewing a virtual subway station (VE) in a helmet mounted display. As they were passively moved within the VE one of three virtual pedestrians randomly approached them from the center (0°), right (+40°) or left (+40°). When present, phone message conditions were delivered at obstacle movement onset and presented either as (1) text messages on the screen of a virtual phone or (2) audio messages delivered through earphones. Participants were instructed to press a joystick button as soon as they detected the approaching virtual pedestrian and to report the message content at the end of the trials.ResultsYoung and older participants showed delayed obstacle detection times with vs. without text messages. Older adults further showed reduced accuracy of message report for texts compared to audio messages. In both groups, audio messages yielded no difference in obstacle detection time or accuracy of message report compared to the no message condition.ConclusionsFindings indicate that text messages prolong the detection of approaching pedestrians, suggesting that they compromise safe ambulation in community environments. Older adults, who show larger deteriorations on the obstacle detection and message deciphering tasks, may be at even greater risk of collision. Audio messages could be a safer alternative for on-the-go communication.

Perceptual evaluation of maneuvering motion illusion for virtual pedestrians

Crowd simulations span a wide spectrum of application domains, most notably video games, evacuation scenarios, and the movie industry. However, it is not obligatory that all virtual populace applications have the primary objective of realistic simulation. In most instances, it is necessary and sufficient that viewers perceive the crowd as plausible. Even for a crowd consisting of agents navigating on linear trajectories without any maneuvers, visual motion illusion elicited by these trajectories might appear to be a natural consequence, causing them to be perceived as wriggling rather than straight. In this respect, we evaluate in this study whether simulated 3D human agents walking with constant, collision-free velocities, induce such a maneuvering motion illusion, aiming toward an efficient real-time crowd simulation. For this purpose, we recorded videos of virtual human crowds with different parameter combinations, such as the agent walking speed, crowd density, camera tilt angle, and camera distance. These videos were watched by human subjects who were instructed to mark the virtual agents who they thought had changed their gait directions. The analyzed results revealed that participants claimed the presence of maneuvering virtual agents in the videos, even though there were none in any of them. Spatial grouping of the markings highlighted that the participants mainly focused on the central area of the simulation environment, and spatiotemporal analysis of the click data also showed stronger evidence to such an illusion (see accompanying video). Furthermore, we found that all of the referred parameters have statistically significant main effects on the number of marked agents per watched video.

Spatial memory enhances the evacuation efficiency of virtual pedestrians under poor visibility condition**Project supported by the Research Grants Council of the Hong Kong Special Administrative Region, China (Grant No. 11203615).

Spatial memory is a critical navigation support tool for disoriented evacuees during evacuation under adverse environmental conditions such as dark or smoky conditions. Owing to the complexity of memory, it is challenging to understand the effect of spatial memory on pedestrian evacuation quantitatively. In this study, we propose a simple method to quantitatively represent the evacueeʼs spatial memory about the emergency exit, model the evacuation of pedestrians under the guidance of the spatial memory, and investigate the effect of the evacueeʼs spatial memory on the evacuation from theoretical and physical perspectives. The result shows that (i) a good memory can significantly assist the evacuation of pedestrians under poor visibility conditions, and the evacuation can always succeed when the degree of the memory exceeds a threshold (); (ii) the effect of memory is superior to that of “follow-the-crowd” under the same environmental conditions; (iii) in the case of multiple exits, the difference in the degree of the memory between evacuees has a significant effect (the greater the difference, the faster the evacuation) for the evacuation under poor visibility conditions. Our study provides a new quantitative insight into the effect of spatial memory on crowd evacuation under poor visibility conditions.

Driving at night with a cataract: Risk homeostasis?

Driving with a cataract can be dangerous, especially at night when road lighting and automotive lighting produce glare. Disability glare alters visual performance, while discomfort glare contributes to restricted mobility, with drivers avoiding driving at night. The present study was focused on the visual effects of an early cataract, and aimed at comparing three driving performance indexes at night, under glare conditions, with and without a simulated cataract. Two indexes directly referred to road safety, while a measure could be related to behavioral adaptation.Using a driving simulator, twenty-six participants were asked to drive in simulated night-time conditions, under controlled photometric conditions where the adaptation luminance and the glare level were consistent with a two-lane rural road at night with oncoming traffic. The visual effects of a bilateral cataract were simulated using goggles which were in the range of an early cataract in terms of light scattering, light transmission, visual acuity and contrast sensitivity loss. Participants were asked to avoid virtual pedestrians on the road, both with and without a simulated cataract. Three performance indexes were considered: the rate of pedestrian crashes, the distance to the pedestrian when the participant avoided the crash, and the mean speed, which allowed to control for a possible behavioral adaptation to the reduced visual performance. For a better understanding of the visual functions responsible for the degraded driving behavior, contrast sensitivity and time-to-collision performance were also measured in glare conditions.While simulated cataract resulted in slightly slower speeds, poorer driving performance was observed with the goggles than without, with more pedestrians being hit and shorter stopping distances. Time-to-collision estimates at 90 km/h were found to be predictive of stopping distances with a simulated cataract, while contrast sensitivity in glare conditions at 13 cycles per degree was found to be associated with the occurrence of a crash with cataract.The decrease in speed with a simulated cataract was real but ineffective in terms of driving safety, which suggests that the behavioral adaptation to the degraded visual performance was insufficient. The precise impact of a cataract on driving abilities remains to be further studied, to provide scientific knowledge to help practitioners determine the moment when the individuals should forego driving.

Crowd Data Visualization and Simulation

Simulating large and complex urban environments for real-time simulation poses multiple challenges which are not trivial to solve, for instance, gathering information of geometry for cities to be displayed, management of large quantities of data and fetching efficiently the information needed for the simulation and visualization of virtual pedestrians are just some examples of the challenges we face.In this work we present the results we have achieved so far using a system that addresses problems such as data visualization using level of detail techniques, GPU processing for efficient computation and the integration of real data into the simulation such as GPS data and real 3D cities that are generated using tools such as WRLD-3D plug-in. This method has been proven powerful enough to handle large crowds composed by thousands of agents in urban environments at interactive frame rates (at least 30 frames per second or 33.33 ms per frame) without compromising the visual quality and the complexity of the behaviors by using publicly real world data such as city topography and pedestrian GPS data.