Anaglyph glasses, which cover the eyes with filters of different color, provide a straightforward way to present stereoscopic stimuli without specialized display or projection hardware. We introduce a 3D-printable anaglyph add-on for the Pupil Labs Neon eye tracker. We compare this setup to high-resolution stationary eye tracking combined with a cold-mirror stereoscope or shutter glasses. We use binocular rivalry, where distinct stimuli are presented to either eye inducing alternating percepts, as testbed. In N=11 observers, we measured how well the behaviorally reported percept could be decoded from eye-tracking data, specifically from the optokinetic nystagmus (OKN). OKN slow phases were of shorter duration and smaller gain with the Pupil Labs anaglyph add-on compared to the setups using stationary eye tracking. Importantly, however, decoding quality was indistinguishable between setups. This demonstrates that our setup presents a viable alternative for dichoptic and stereoscopic presentation, especially if mobility and cost efficiency are desirable.

The decision of whether to cross a road or wait for a car to pass, humans make frequently and effortlessly. Recently, the application of drift-diffusion models (DDMs) on pedestrians’ decision-making has proven useful in modelling crossing behaviour in pedestrian-vehicle interactions. These models consider binary decision-making as an incremental accumulation of noisy evidence over time until one of two choice thresholds (to cross or not) is reached. One open question is whether the assumption of a kinematics-dependent drift-diffusion process, which was made in previous pedestrian crossing DDMs, is justified, with DDM-parameters varying over time according to the developing traffic situation. It is currently unknown whether kinematics-dependent DDMs provide a better model fit than conventional DDMs, which are fitted per condition. Furthermore, previous DDMs have not considered reaction times for the not-crossing option. We address these issues by a novel experimental design combined with modelling. Experimentally, we use a 2-alternative-forced-choice paradigm, where participants view videos of approaching cars from a pedestrian’s perspective and respond whether they want to cross before the car or to wait until the car has passed. Using these data, we perform thorough model comparison between kinematics-dependent and condition-wise fitted DDMs. Our results demonstrate that condition-wise fitted DDMs can show better model fits than kinematics-dependent DDMs as reflected in the mean-squared-errors. The condition-wise fitted models need considerably more parameters, but in some cases still outperform kinematics-dependent DDMs in measures that penalize the parameter number (e.g., Akaike information criterion). Introducing a starting point bias provides support for the novel hypothesis of rapid early evidence build-up from the initial view of the vehicle distance. The drift rates obtained for the condition-wise fitted models align with the assumptions in the kinematics-dependent models, confirming that pedestrians’ decision processes are kinematics-dependent. However, the partial preference for condition-wise fitted models in the model selection suggests that the correct form of kinematics-dependence has not yet been identified for all DDM-parameters, indicating room for improvement of current pedestrian crossing DDMs. Developing more accurate models of human cognitive processes will likely facilitate autonomous vehicles to understand pedestrians’ intentions as well as to show unambiguous human-like behaviour in future traffic interactions with humans.

Using a map in an unfamiliar environment requires identifying correspondences between elements of the map's allocentric representation and elements in egocentric views. Aligning the map with the environment can be challenging. Virtual reality (VR) allows learning about unfamiliar environments in a sequence of egocentric views that correspond closely to the perspectives and views that are experienced in the actual environment. We compared three methods to prepare for localization and navigation tasks performed by teleoperating a robot in an office building: studying a floor plan of the building and two forms of VR exploration. One group of participants studied a building plan, a second group explored a faithful VR reconstruction of the building from a normal-sized avatar's perspective, and a third group explored the VR from a giant-sized avatar's perspective. All methods contained marked checkpoints. The subsequent tasks were identical for all groups. The self-localization task required indication of the approximate location of the robot in the environment. The navigation task required navigation between checkpoints. Participants took less time to learn with the giant VR perspective and with the floorplan than with the normal VR perspective. Both VR learning methods significantly outperformed the floorplan in the orientation task. Navigation was performed quicker after learning in the giant perspective compared to the normal perspective and the building plan. We conclude that the normal perspective and especially the giant perspective in VR are viable options for preparing for teleoperation in unfamiliar environments when a virtual model of the environment is available.

Social robots are often applied in recreational contexts to improve the experience of using technical systems, but they are also increasingly used for therapeutic purposes. In this study, we compared how patients with Autism Spectrum Disorder (ASD) interact with a social robot and a human actor. We examined the gaze behavior of nine ASD patients and 15 matched controls using a mobile eye-tracker. Participants performed a task in which they were required to follow the gaze of a robot or human actor. Our results show that ASD patients preferentially maintain eye contact during interaction with the social robot as compared to the human actor.

Under natural viewing conditions, human observers shift their gaze to allocate processing resources to subsets of the visual input. Many computational models have aimed at predicting such voluntary attentional shifts. Although the importance of high level stimulus properties (higher order statistics, semantics) stands undisputed, most models are based on low-level features of the input alone. In this study we recorded eye-movements of human observers while they viewed photographs of natural scenes. About two thirds of the stimuli contained at least one person. We demonstrate that a combined model of face detection and low-level saliency clearly outperforms a low-level model in predicting locations humans fixate. This is reflected in our finding fact that observes, even when not instructed to look for anything particular, fixate on a face with a probability of over 80% within their first two fixations (500ms). Remarkably, the model's predictive performance in images that do not contain faces is not impaired by spurious face detector responses, which is suggestive of a bottom-up mechanism for face detection. In summary, we provide a novel computational approach which combines high level object knowledge (in our case: face locations) with low-level features to successfully predict the allocation of attentional resources.

We present a method to learn models of human heads for the purpose of detection from different viewing angles. We focus on a model where objects are represented as constellations of rigid features (parts). Variability is represented by a joint probability density function (PDF) on the shape of the constellation. In the first stage, the method automatically identifies distinctive features in the training set using an interest operator followed by vector quantization. The set of model parameters, including the shape PDF, is then learned using expectation maximization. Experiments show good generalization performance to novel viewpoints and unseen faces. Performance is above 90% correct with less than 1 s computation time per image.