To Örebro University

There is an increased interest in using social robots to assist older adults during their daily life activities. As social robots are designed to interact with older users, it becomes relevant to study these interactions under the lens of social cognition. Gaze following, the social ability to infer where other people are looking at, deteriorates with older age. Therefore, the referential gaze from robots might not be an effective social cue to indicate spatial locations to older users. In this study, we explored the performance of older adults, middle-aged adults, and younger controls in a task assisted by the referential gaze of a Pepper robot. We examined age-related differences in task performance, and in self-reported social perception of the robot. Our main findings show that referential gaze from a robot benefited task performance, although the magnitude of this facilitation was lower for older participants. Moreover, perceived anthropomorphism of the robot varied less as a result of its referential gaze in older adults. This research supports that social robots, even if limited in their gazing capabilities, can be effectively perceived as social entities. Additionally, this research suggests that robotic social cues, usually validated with young participants, might be less optimal signs for older adults.

Supplementary Information: The online version contains supplementary material available at 10.1007/s12369-022-00926-6.

Mentalizing, where humans infer the mental states of others, facilitates understanding and interaction in social situations. Humans also tend to adopt mentalizing strategies when interacting with robotic agents. There is an ongoing debate about how inferred mental states affect gaze following, a key component of joint attention. Although the gaze from a robot induces gaze following, the impact of mental state attribution on robotic gaze following remains unclear. To address this question, we asked forty-nine young adults to perform a gaze cueing task during which mental state attribution was manipulated as follows. Participants sat facing a robot that turned its head to the screen at its left or right. Their task was to respond to targets that appeared either at the screen the robot gazed at or at the other screen. At the baseline, the robot was positioned so that participants would perceive it as being able to see the screens. We expected faster response times to targets at the screen the robot gazed at than targets at the non-gazed screen (i.e., gaze cueing effect). In the experimental condition, the robot's line of sight was occluded by a physical barrier such that participants would perceive it as unable to see the screens. Our results revealed gaze cueing effects in both conditions although the effect was reduced in the occluded condition compared to the baseline. These results add to the expanding fields of social cognition and human-robot interaction by suggesting that mentalizing has an impact on robotic gaze following.

A significant challenge in service robots is the semantic understanding of their surrounding areas. Traditional approaches addressed this problem by segmenting the environment into regions corresponding to full rooms that are assigned labels consistent with human perception, e.g. office or kitchen. However, different areas inside the same room can be used in different ways: Could the table and the chair in my kitchen become my office ? What is the category of that area now? office or kitchen? To adapt to these circumstances we propose a new paradigm where we intentionally relax the resulting labeling of place classifiers by allowing confusions, and by avoiding further filtering leading to clean full room classifications. Our hypothesis is that confusions can be beneficial to a service robot and, therefore, they can be kept and better exploited. Our approach creates a subdivision of the environment into different regions by maintaining the confusions which are due to the scene appearance or to the distribution of objects. In this paper, we present a proof of concept implemented in simulated and real scenarios, that improves efficiency in the robotic task of searching for objects by exploiting the confusions in place classifications.

Cardiac magnetic resonance (CMR) imaging is widely used to visualise cardiac motion and diagnose heart disease. However, standard CMR imaging requires patients to lie still in a confined space inside a loud machine for 40-60 min, which increases patient discomfort. In addition, shorter scan times decrease either or both the temporal and spatial resolutions of cardiac motion, and thus, the diagnostic accuracy of the procedure. Of these, we focus on reduced temporal resolution and propose a neural network called FusionNet to obtain four-dimensional (4D) cardiac motion with high temporal resolution from CMR images captured in a short period of time. The model estimates intermediate 3D heart shapes based on adjacent shapes. The results of an experimental evaluation of the proposed FusionNet model showed that it achieved a performance of over 0.897 in terms of the Dice coefficient, confirming that it can recover shapes more precisely than existing methods. This code is available at: https://github.com/smiyauchi199/FusionNet.git.

The ability to follow the gaze of others deteriorates with age. This decline is typically tested with gaze cueing tasks, in which the time it takes to respond to targets on a screen is faster when they are preceded by a facial cue looking in the direction of the target (i.e., gaze cueing effect). It is unclear whether age-related differences in this effect occur with gaze cues other than the eyes, such as head orientation, and how these vary in function of the cue-target timing. Based on the perceived usefulness of social robots to assist older adults, we asked older and young adults to perform a gaze cueing task with the head of a NAO robot as the central cue. Crucially, the head was viewed from the back, and so its eye gaze was conveyed. In a control condition, the head was static and faced away from the participant. The stimulus onset asynchrony (SOA) between cue and target was 340 ms or 1000 ms. Both age groups showed a gaze cueing effect at both SOAs. Older participants showed a reduced facilitation effect (i.e., faster on congruent gazing trials than on neutral trials) at the 340-ms SOA compared to the 1000-ms SOA, and no differences between incongruent trials and neutral trials at the 340-ms SOA. Our results show that a robot with non-visible eyes can elicit gaze cueing effects. Age-related differences in the other effects are discussed regarding differences in processing time.

This letter addresses the problem of human activity recognition (HAR) of people wearing inertial sensors using data from the UCI-HAR dataset. We propose a light residual network, which obtains an F1-Score of 97.6% that outperforms previous works, while drastically reducing the number of parameters by a factor of 15, and thus the training complexity. In addition, we propose a new benchmark based on leave-one (person)-out cross-validation to standardize and unify future classifications on the same dataset, and to increase reliability and fairness in the comparisons.

Autonomous systems in the road transportation network require intelligent mechanisms that cope with uncertainty to foresee the future. In this paper, we propose a multi-stage probabilistic approach for trajectory forecasting: trajectory transformation to displacement space, clustering of displacement time series, trajectory proposals, and ranking proposals. We introduce a new deep feature clustering method, underlying self-conditioned GAN, which copes better with distribution shifts than traditional methods. Additionally, we propose novel distance-based ranking proposals to assign probabilities to the generated trajectories that are more efficient yet accurate than an auxiliary neural network. The overall system surpasses context-free deep generative models in human and road agents trajectory data while performing similarly to point estimators when comparing the most probable trajectory.

Robots can direct human attention using their eyes. However, it remains unclear whether it is the gaze or the low-level motion of the head rotation that drives attention. We isolated these components in a non-predictive gaze cueing task with a robot to explore how limited robotic signals orient attention. In each trial, the head of a NAO robot turned towards the left or right. To isolate the direction of rotation from its gaze, NAO was presented frontally and backward along blocks. Participants responded faster to targets on the gazed-at site, even when the eyes of the robot were not visible and the direction of rotation was opposed to that of the frontal condition. Our results showed that low-level motion did not orient attention, but the gaze direction of the robot did. These findings suggest that the robotic gaze is perceived as a social signal, similar to human gaze.

Autonomous systems, that need to operate in human environments and interact with the users, rely on understanding and anticipating human activity and motion. Among the many factors which influence human motion, semantic attributes, such as the roles and ongoing activities of the detected people, provide a powerful cue on their future motion, actions, and intentions. In this work we adapt several popular deep learning models for trajectory prediction with labels corresponding to the roles of the people. To this end we use the novel THOR-Magni dataset, which captures human activity in industrial settings and includes the relevant semantic labels for people who navigate complex environments, interact with objects and robots, work alone and in groups. In qualitative and quantitative experiments we show that the role-conditioned LSTM, Transformer, GAN and VAE methods can effectively incorporate the semantic categories, better capture the underlying input distribution and therefore produce more accurate motion predictions in terms of Top-K ADE/FDE and log-likelihood metrics.

We present a system designed to monitor the well-being of older adults during their daily activities. To automatically detect and classify their emotional state, we collect physiological data through a wearable medical sensor. Ground truth data are obtained using a simple smartphone app that provides ecological momentary assessment (EMA), a method for repeatedly sampling people's current experiences in real time in their natural environments. We are making the resulting dataset publicly available as a benchmark for future comparisons and methods. We are evaluating two feature selection methods to improve classification performance and proposing a feature set that augments and contrasts domain expert knowledge based on time-analysis features. The results demonstrate an improvement in classification accuracy when using the proposed feature selection methods. Furthermore, the feature set we present is better suited for predicting emotional states in a leave-one-day-out experimental setup, as it identifies more patterns.