Stefan-Daniel Suvei

I would like to get close to you: Making robot personal space invasion less intrusive with a social gaze cue

How can a social robot get physically close to the people it needs to interact with? We investigated the effect of a social gaze cue by a human-sized mobile robot on the effects of personal space invasion by that robot. In our 2\(\,\times \,\)2 between-subject experiment, our robot would approach our participants (n = 83), with/without personal space invasion, and with/without a social gaze cue. With a questionnaire, we measured subjective perception of warmth, competence, and comfort after such an interaction. In addition, we used on-board sensors and a tracking system to measure the dynamics of social positioning behavior. While we did find significant differences in the social positioning dynamics of the participants, no such effect was found upon quantitative analysis of perception of the robot. In a subsequent inductive analysis we further investigated these results, our findings suggesting that the social cue did play a role for the participants – particularly related to their perceived safety.

Stereo and LIDAR Fusion based Detection of Humans and Other Obstacles in Farming Scenarios.

In this paper we propose a fusion method which uses the depth information acquired from a LIDAR sensor to guide a block matching stereo algorithm. The resulting fused point clouds are then used for obstacle detection, either by processing the raw data and clustering the protruding objects in the scene, or by applying a Convolutional Neural Network on the 3D points and labeling them into classes. The performance of the proposed method is evaluated by carrying out a series of experiments on different data sets obtained from the SAFE robotic platform. The results show that the fusion algorithm significantly improves the F1 detection score of the trained networks.

Timing of multimodal robot behaviors during human-robot collaboration

In this paper, we address issues of timing between robot behaviors in multimodal human-robot interaction. In particular, we study what effects sequential order and simultaneity of robot arm and body movement and verbal behavior have on the fluency of interactions. In a study with the Care-O-bot, a large service robot, in a medical measurement scenario, we compare the timing of the robots behaviors in three between-subject conditions. The results show that the relative timing of robot behaviors has significant effects on the number of problems participants encounter, and that the robots verbal output plays a special role because participants carry their expectations from human verbal interaction into the interactions with robots.

Applying Peg-in-Hole Actions with a Service Robot

A general requirement for any service robot is to be flexible and capable of processing uncertainties, thus making it adaptable for multiple tasks. As a result, learning the appropriate action parameters for a specific action is a crucial task. The method presented in this paper is an incremental statistical learning method that takes into consideration the uncertainties and the contact forces to find the optimal parameter sets. The method is inspired by solutions available in industrial robotics and it uses a dynamic simulator and Kernel Density Estimation in order to find the parameter sets that lead to a successful Peg-in-Hole action. The solution obtained in the simulation is successfully tested on a real service robot.

Between legibility and contact: The role of gaze in robot approach

In this paper, we explore experimentally the possible tradeoff between gaze to the user and gaze to the path in robot approach. While some previous work indicates that gaze towards the user increases perceived safety because the user feels recognized, other work indicates that it is legibility of the robots actions that put users at ease. If the robot does not drive up to the person in a straight line directly, the robot can either continuously look at the person and thus maintain eye contact, or indicate its path through its gaze behavior, increasing legibility. In an experiment with N=36 participants, we tested the tradeoff between legibility and eye contact. The behavioral results show that users are significantly more at ease with the robot that gazes at them than with the robot that looks where it is going, measured by the number of instances of glances away from the robot. Likewise, the participants rate the robot that looks at them continuously as more intelligent and more cooperative. Thus, participants value mutual gaze higher than legibility.

Stereo and Active-Sensor Data Fusion for Improved Stereo Block Matching

This paper proposes an algorithm which uses the depth information acquired from an active sensor as guidance for a block matching stereo algorithm. In the proposed implementation, the disparity search interval used for the block matching is reduced around the depth values obtained from the active sensor, which leads to an improved matching quality and denser disparity maps and point clouds. The performance of the proposed method is evaluated by carrying out a series of experiments on 3 different data sets obtained from different robotic systems. We demonstrate with experimental results that the disparity estimation is improved and denser disparity maps are generated.