Robot Perception Research Articles

Dense Depth-Map Estimation Based on Fusion of Event Camera and Sparse LiDAR

Depth-map estimation reflects the geometry of the visible surface in the environment directly and plays an important role in perception and decision for intelligent robots. However, sparse LiDAR only provides low-resolution depth information, which is a huge challenge for accurate sensing algorithms. To address this problem, this article proposes a novel fusion framework to generate dense depth-map based on event camera and sparse LiDAR. The approach uses the geometric information provided by the point cloud as prior knowledge and clusters point cloud data by an improved density clustering algorithm. Combined with the 3-D surface model of each cluster, the approach can provide 3-D reconstructions of the coordinate points of events and further obtain the dense-depth map by depth expansion and hole filling. Finally, we deploy our approach in MVSEC datasets and real-world applications. Experimental results show that, compared with other approaches, our approach can obtain more accurate depth information.

Frontline Employees’ Attitude Toward Embodied Social Robots In Customer Service: An Integrative Framework And Empirical Test

Embodied social robots-robots providing services for and in cocreation with consumers-are expected to profoundly change the way services are delivered. Yet, their integration in customer service poses a challenge: their adoption by frontline employees (FLEs). Accordingly, this study aims to examine FLEs’ attitude toward embodied social robots and to uncover its antecedents. This work presents an integrative framework which builds upon the technology acceptance model and examines the influence of potential factors on FLEs’ attitude toward embodied social robots. An online survey among 165 FLEs is used to test the integrative framework. Despite the growing knowledge regarding customers’ perceptions of (embodied social) robots, the perspective of FLEs is under-investigated while crucial to foster FLEs’ acceptance of such robots. This research concludes with several strategies service providers can implement to possibly enhance FLEs’ attitude toward embodied social robots, and thus, to support their adoption by FLEs.

Vision-based Moving Target Tracking of Certain Target for Quadruped Robots

Due to the high flexibility of quadruped robots compared with some traditional robots, it has become an important branch in the field of mobile robot research. Target detection and tracking technology is important for the environment perception and autonomous motion tasks of quadruped robots. To achieve effective tracking of a specific target, the premise is that the target is accurately detected. In this article, a target detection algorithm based on deep learning for quadruped robot person tracking is proposed. After the targets are detected, a simple but accurate selection strategy is used to extract the certain target to be tracked from other targets. The visual distance and angle information from the target is used to realize the autonomous tracking. In the part of experiment, we conducted experiments based on the changeable outdoor environment, which included interference factors such as light changes, people crossing, obstruction by obstacles. During the experiments, the SDUQuad-48 quadruped robot can simultaneously realize the stable tracking of a specific moving target based on our algorithm with a speed of 0.8 m/s.

Frontline Employees’ Attitude Toward Embodied Social Robots In Customer Service: An Integrative Framework And Empirical Test

Embodied social robots-robots providing services for and in cocreation with consumers-are expected to profoundly change the way services are delivered. Yet, their integration in customer service poses a challenge: their adoption by frontline employees (FLEs). Accordingly, this study aims to examine FLEs’ attitude toward embodied social robots and to uncover its antecedents. This work presents an integrative framework which builds upon the technology acceptance model and examines the influence of potential factors on FLEs’ attitude toward embodied social robots. An online survey among 165 FLEs is used to test the integrative framework. Despite the growing knowledge regarding customers’ perceptions of (embodied social) robots, the perspective of FLEs is under-investigated while crucial to foster FLEs’ acceptance of such robots. This research concludes with several strategies service providers can implement to possibly enhance FLEs’ attitude toward embodied social robots, and thus, to support their adoption by FLEs.

Data augmentation methods for semantic segmentation-based mobile robot perception system

Data augmentation has become a standard technique for increasing deep learning models? accuracy and robustness. Different pixel intensity modifications, image transformations, and noise additions represent the most utilized data augmentation methods. In this paper, a comprehensive evaluation of data augmentation techniques for mobile robot perception system is performed. The perception system based on a deep learning model for semantic segmentation is augmented by 17 techniques to obtain better generalization characteristics during the training process. The deep learning model is trained and tested on a custom dataset and utilized in real-time scenarios. The experimental results show the increment of 6.2 in mIoU (mean Intersection over Union) for the best combination of data augmentation strategies.

HiPE: Hierarchical Initialization for Pose Graphs

Pose graph optimization is a non-convex optimization problem encountered in many areas of robotics perception. Its convergence to an accurate solution is conditioned by two factors: the non-linearity of the cost function in use and the initial configuration of the pose variables. In this paper, we present HiPE, a novel hierarchical algorithm for pose graph initialization. Our approach exploits a coarse-grained graph that encodes an abstract representation of the problem geometry. We construct this graph by combining maximum likelihood estimates coming from local regions of the input. By leveraging the sparsity of this representation, we can initialize the pose graph in a non-linear fashion, without computational overhead compared to existing methods. The resulting initial guess can effectively bootstrap the fine-grained optimization that is used to obtain the final solution. In addition, we perform an empirical analysis on the impact of different cost functions on the final estimate. Our experimental evaluation shows that the usage of HiPE leads to a more efficient and robust optimization process, comparing favorably with state-of-the-art methods.

Targetless Extrinsic Calibration of Stereo, Thermal, and Laser Sensors in Structured Environments

2D and 3D sensor extrinsic calibration is the key prerequisite for multi-sensor based robot perception and localization. However, such calibration is challenging due to the variety of sensor modalities and the requirement of special calibration targets and human intervention. In this paper, we demonstrate a new targetless cross-modal calibration system focusing on the extrinsic transformations among stereo cameras, thermal cameras, and laser sensors. Specifically, the calibration between stereo and laser is conducted in 3D space by minimizing the registration error, while the thermal extrinsic to the other two sensors is estimated by optimizing the alignment of the edge features. Due to the low contrast of thermal images, the extracted edges are often noisy, resulting in incorrect edge matching. We introduce the edge alignment optimization on attraction field map to overcome this challenge. Our method requires no dedicated targets and performs the multi-sensor calibration in a single shot without human interaction. Extensive experiments on our collected real-world datasets show that our system can be easily used in structured environments with high extrinsic calibration accuracy. The video demonstration can be found at https://www. youtube.com/watch?v=b1QAVY7hUu8. The code is released to the public at https://github.com/FuTaimeng/auto calibration.

Elephants Don’t Pack Groceries: Robot Task Planning for Low Entropy Belief States

Recent advances in computational perception have significantly improved the ability of autonomous robots to perform state estimation with low entropy. Such advances motivate a reconsideration of robot decision-making under uncertainty. Current approaches to solving sequential decision-making problems model states as inhabiting the extremes of the perceptual entropy spectrum. As such, these methods are either incapable of overcoming perceptual errors or asymptotically inefficient in solving problems with low perceptual entropy. With low entropy perception in mind, we aim to explore a happier medium that balances computational efficiency with the forms of uncertainty we now observe from modern robot perception. We propose an approach for efficient task planning for goal-directed robot reasoning. Our approach combines belief space representation with the fast, goal-directed features of classical planning to efficiently plan for low entropy goal-directed reasoning tasks. We compare our approach with current classical planning and belief space planning approaches by solving low entropy goal-directed grocery packing tasks in simulation. Our approach outperforms these approaches in planning time, execution time, and task success rate in our simulation experiments. We also demonstrate our approach on a real world grocery packing task with physical robot.

Impact of nonverbal robot behaviour on human teachers’ perceptions of a learner robot

AbstractHow do we perceive robots practising a task that we have taught them? While learning, human trainees usually provide nonverbal cues that reveal their level of understanding and interest in the task. Similarly, nonverbal social cues of trainee robots that can be interpreted naturally by humans can enhance robot learning. In this article, we investigated a scenario in which a robot is practising a physical task in front of the human teachers (i.e., participants), who were asked to assume that they had previously taught the robot to perform that task. Through an online experiment with 167 participants, we examined the effects of different gaze patterns and arm movements with multiple speeds and various kinds of pauses on human teachers’ perception of different attributes of the robot. We found that the perception of a trainee robot’s attributes (e.g., confidence and eagerness to learn) can be systematically affected by its behaviours. Findings of this study can inform designing more successful nonverbal social interactions for intelligent robots.

Experiences and perceptions of using a low-cost pet robot for older adults and people with dementia

Pet robots are a practicable substitute for animal-assisted therapy. They have been shown to have positive impacts on older adults, including people with dementia, such as providing companionship and facilitating social interaction. However, the issue of affordability can hinder equal access to such technology. The purpose of our study was to understand the perceptions and experiences of using a low-cost, commercially available pet robot with older adults and people with dementia. We used a novel methodology of analysing a large volume of user reviews that were collected from 15 consumer websites. A total of 1,327 user reviews that met our pre-specified inclusion criteria were included. Descriptive statistics was applied to characterise demographic data, and inductive qualitative content analysis was used to identify themes in the textual data. Most reviews were obtained from consumer sites in the United States, and most reviewers were family members of the users (i.e., older adults and people with dementia). We found that circumstantial reasons, such the inability to own live animals, prompted reviewers to purchase the pet robot. Most reviewers had positive perceptions of the pet robot, and described various activities that users engaged in with it. Impacts of using the pet robot, such as positive emotions, were also described. Finally, experiences about practical aspects of its use, such as durability and hygiene, were discussed. Overall, this study provides useful knowledge that can help researchers, robot developers and clinicians understand the viability of using low-cost pet robots to benefit older adults, including people with dementia.

A growth mindset about human minds promotes positive responses to intelligent technology

Perceiving minds in technology agents, for example, robots designed with artificial intelligence (AI), is common and crucial in modern life. However, past studies have revealed that robots with a high level of minds elicit polarized responses. From a human-robot interaction perspective, we proposed that people’ responses to robots, in part, originate from differences between fixed and growth mindsets about human minds—the beliefs regarding whether humans' mental capacities are fixed or incremental. We conducted five studies to test this assumption. A growth mindset about human minds was associated with or led to lower levels of negative feelings about robots (Study 1), more perceptions of robots as allies versus enemies (Study 2), more support for robotic research (Studies 3 and 4), and greater willingness to interact with robots (Study 5). Furthermore, the effect of a growth mindset about human minds on favorable responses to robots was more pronounced when robots were perceived as having high (versus low) levels of minds (Studies 3–5) and mediated by decreased concerns about robots (Study 5). By emphasizing the nuanced role of mindset beliefs about human minds in responses to intelligent technology, this research provides not only a new perspective on research into minds but also important implications for human-technology relationships.

Sensing Deformation in Vacuum Driven Foam-Based Actuator via Inductive Method.

Perception in soft robotics is crucial to allow a safe interaction to effectively explore the environment. Despite the inherent capabilities of soft materials, embedding reliable sensing in soft actuators or robots could introduce constraints in the overall design (e.g., loss of deformability, undesired trajectories, etc.) or reduce their compliant characteristics. Consequently, an adequate stiffness for both sensor and actuator becomes a crucial design parameter. In particular, for sensing the deformation related to actuation motion, sensing and actuating strategies must work in full mechanical synergy. In this view, an inductive sensing solution is presented, exploiting open-cell foam and a copper (Cu) wire in an Inductive Foam Sensor (IFS). Due to entangled air cells high deformability is enabled upon vacuum pressure, and proprioceptive information is provided. The IFS is then successfully integrated into the earlier developed Ultralight Hybrid Pneumatic Artificial Muscle (UH-PAM), which encases an elastomeric bellow skin and plastic rings. Such sensorized UH-PAM (SUH-PAM) is capable of a high contraction ratio (54% upon -80kPa), while the inductive sensing shows a high sensitivity of 0.01031/1% and a hysteresis of 5.35%, with an average error of 1.85%, respectively. In order to implement a robust feedback control system, an adaptable proportional sliding mode control is presented. As a result, the SUH-PAM motion can be controlled to the mm-scale, with an RMSE of 0.925 mm, and high robustness against disturbances is demonstrated.

On sharing and caring: Investigating the effects of a robot's self-disclosure and question- asking on children's robot perceptions and child-robot relationship formation

Although scholars have focused on the role of self-disclosure in the context of child-robot interaction and relationship formation, little is known as to how the effects of a robot's self-disclosure vary by the information the robot shares. Moreover, the influence of a robot's question-asking on children's perception of, and relationship formation with the robot remains understudied. We therefore investigated experimentally, with a 2 × 2 between-subjects design, how children perceive and relate to a social robot when the robot engages in (personal) self-disclosure versus (factual) self-description and when the robot does or does not ask them questions. We collected self-report data from 293 children aged 7 to 10, who interacted with the Nao robot in a science museum. The robot's question-asking increased children's trust in the robot as well as their belief in its cognitive perspective-taking abilities. Self-disclosure, in contrast, decreased children's perception of the robot's capacity to adopt their affective perspective. Children's consideration of the robot as a social actor and a potential friend did not differ across conditions. A post-hoc analysis showed that the effect of question-asking on trust was mediated by children's perception of the robot's cognitive perspective-taking abilities.

Appearance is not everything - Preferred feature combinations for care robots

To promote acceptance and use of social robots, the application of lifelike design features has been established as a promising design principle. The majority of research to date emphasizes a high degree of anthropomorphism with respect to robot appearance and communication to support positive perceptions of robots. However, these features are often examined separately. As a result, there has been little evidence of possible interaction effects when implementing different design principles with a particular focus on matched features. An online survey with two different samples (sample 1: random; sample 2: healthcare professionals and trainees) is reported that investigate whether certain feature combinations are preferred for care robots. The attributes appearance (human-like, animal-like, technical), size (small, large), and communication (speech, sound, text) were combined. Results suggest that communication is the key determinant of robot preference. With regard to feature combinations the survey showed an overall preference for anthropomorphic features. Furthermore, the occurrence of matched features led to an additional improvement of preference ratings. Both samples provide similar results with high correlations. The obtained findings can therefore serve as a basis for the implementation of concrete design principles for social robotics in healthcare.

Implementation and Application of Fall Detection on Mobile Perception Robot

With the rapid increase in the number of people living in the elderly population, reducing and dealing with the problem of falls in the elderly has become the focus of research for decades. It is impossible to completely eliminate falls in daily life and activities. Detecting a fall in time can protect the elderly from injury as much as possible. This article uses the Turtlebot robot and the ROS robot operating system, combined with simultaneous positioning and map construction technology, Monte Carlo positioning, A* path planning, dynamic window method, and indoor map navigation. The YOLO network is trained using the stance and fall data sets, and the YOLOv4 target detection algorithm is combined with the robot perception algorithm to finally achieve fall detection on the turtlebot robot, and use the average precision, precision, recall and other indicators to measure.

Do People Trust in Robot-Assisted Surgery? Evidence from Europe.

(1) Background: The goal of the paper was to establish the factors that influence how people feel about having a medical operation performed on them by a robot. (2) Methods: Data were obtained from a 2017 Flash Eurobarometer (number 460) of the European Commission with 27,901 citizens aged 15 years and over in the 28 countries of the European Union. Logistic regression (odds ratios, OR) to model the predictors of trust in robot-assisted surgery was calculated through motivational factors, using experience and sociodemographic independent variables. (3) Results: The results obtained indicate that, as the experience of using robots increases, the predictive coefficients related to information, attitude, and perception of robots become more negative. Furthermore, sociodemographic variables played an important predictive role. The effect of experience on trust in robots for surgical interventions was greater among men, people between 40 and 54 years old, and those with higher educational levels. (4) Conclusions: The results show that trust in robots goes beyond rational decision-making, since the final decision about whether it should be a robot that performs a complex procedure like a surgical intervention depends almost exclusively on the patient’s wishes.

“My colleague is a robot” – exploring frontline employees' willingness to work with collaborative service robots

PurposeAs service robots increasingly interact with customers at the service encounter, they will inevitably become an integral part of employee's work environment. This research investigates frontline employee's perceptions of collaborative service robots (CSR) and introduces a new framework, willingness to collaborate (WTC), to better understand employee–robot interactions in the workplace.Design/methodology/approachDrawing on appraisal theory, this study employed an exploratory research approach to investigate frontline employees' cognitive appraisal of service robots and their WTC with their nonhuman counterparts in service contexts. Data collection consisted of 36 qualitative problem-centered interviews. Following an iterative thematic analysis, the authors introduce a research framework of frontline employees' WTC with service robots.FindingsFirst, this study demonstrates that the interaction between frontline employees and service robots is a multistage appraisal process based on adoption-related perceptions. Second, it identifies important attributes across three categories (employee, robot and job attributes) that provide a foundation to understand the appraisal of CSRs. Third, it presents four employee personas (supporter, embracer, resister and saboteur) that provide a differentiated perspective of how service employee–robot collaboration may differ.Practical implicationsThe article identifies important factors that enable and restrict frontline service employees' (FSEs’) WTC with robots.Originality/valueThis is the first paper that investigates the appraisal of CSRs from the perspective of frontline employees. The research contributes to the limited research on human–robot collaboration and expands existing technology acceptance models that fall short to explain post-adoptive coping behavior of service employees in response to service robots in the workplace.

Preliminary validation of the European Portuguese version of the Robotic Social Attributes Scale ( RoSAS )

People's perception of social robots is essential in determining their responses and acceptance of this type of agent. Currently, there are few instruments validated for the European Portuguese population that measure the perception of social robots. Our goal was to translate, validate, and evaluate the psychometric properties of the Robotic Social Attributes Scale (RoSAS) to European Portuguese. To achieve this goal, we conducted a validation study using a sample of 185 participants. We measured the temporal validity of the scale (over a 2-week interval) and its divergent and convergent validity using the Portuguese Negative Attitudes toward Robots Scale and the Godspeed scales. Our data analysis resulted in a shortened version of the Portuguese RoSAS with 11 items while retaining the original three-factor structure. The scale presented poor to acceptable levels of temporal reliability. We found a positive correlation between the warmth and competence dimensions. Further validation studies are needed to investigate the psychometric properties of this scale.

When service robots look at themselves in the mirror: An examination of the effects of perceptions of robotic self-recognition

Service robots sharing the same environments with humans are typically designed to have various humanlike features, because perceptions of robots as humanlike make them more acceptable to human users. This study examines one human characteristic that has hitherto been unexplored in studies of humans' perceptions of robots: a self-recognition ability. For (adult) humans, self-recognition is typically uncomplicated and effortless, yet it is of fundamental importance for social interaction, and therefore it was assumed that perceptions of robotic self-recognition abilities would be used for inferences about other robot attributes. A between-subjects experiment, in which a domestic service robot's ability to recognize itself was manipulated (low vs. high), showed that high self-recognition boosted perceptions of the quality of the service delivered by the robot, and that this effect was mediated by perceptions of the robot as having a capacity for learning and having theory of mind.

Versatile and Suitable for Harsh Environments-Capacitive Proximity Perception in Robotics

Robotic systems are becoming increasingly present in every aspect of our daily private as well as working environments. The more these systems enter into our spheres of action, the more they are intended to support humans in dangerous, dull and dirty tasks, and the more flexibility is expected from them. In order to prepare robotic systems for such duties, they need to be able to perceive their environment. In robotic systems, a perception of the environment can either be achieved through internal sensors, i.e., joint angle and torque sensors, or through external sensors, e.g., vision or tactile sensors. While the use of such internal sensors may seem advantageous because no additional system has to be added, the information that can be obtained from them is rather undifferentiated when we want to gather information about the environment. Thus, this information is in most cases augmented by external sensors. Considering exteroception, the most often used technology to analyze the environment is vision. However, due to general deficiencies of the system such as their dependence on lighting conditions and the necessary line of sight, the information accessible through vision alone [1] is not sufficient to accomplish differentiated manipulation tasks [2].