Beatriz Quintino Ferreira

Personalized Assistance for Dressing Users

In this paper, we present an approach for a robot to provide personalized assistance for dressing a user. In particular, given a dressing task, our approach finds a solution involving manipulator motions and also user repositioning requests. Specifically, the solution allows the robot and user to take turns moving in the same space and is cognizant of the user’s limitations. To accomplish this, a vision module monitors the human’s motion, determines if he is following the repositioning requests, and infers mobility limitations when he cannot. The learned constraints are used during future dressing episodes to personalize the repositioning requests. Our contributions include a turn-taking approach to human-robot coordination for the dressing problem and a vision module capable of learning user limitations. After presenting the technical details of our approach, we provide an evaluation with a Baxter manipulator.

A unified approach for hybrid source localization based on ranges and video

This paper presents a hybrid method for single-source localization in wireless sensor networks, fusing noisy range measurements with angular information extracted from video. Although recent works found in the literature explore hybrid schemes, these include several cumbersome assumptions. We develop and test, both numerically and experimentally, a hybrid localization algorithm which surpasses the limitations of previous fusing approaches. The proposed method (FLORIS) is based on a nonconvex least-squares joint formulation, for which a tight convex relaxation is applied to obtain a semidefinite program. Numerical simulations show that FLORIS has comparable performance to state-of-the-art methods, even outperforming them in some scenarios. Real experiments show that FLORIS is feasible in practical application scenarios, achieving very good accuracy and robustness. Importantly, coverage requirements for the infrastructure in a given area are more flexible than resorting to a single type of sensor, which may simplify practical deployments.

Developing and Evaluating Two Gestural-Based Virtual Environment Navigation Methods for Large Displays

In this paper we present two methods to navigate in virtual environments displayed in a large display using gestures detected by a depth sensor. We describe the rationale behind the development of these methods and a user study to compare their usability performed with the collaboration of 17 participants. The results suggest the users have a better performance and prefer one of them, while considering both as suitable and natural navigation methods.

VOLUMNECT: measuring volumes with Kinect

This article presents a solution to volume measurement object packing using 3D cameras (such as the Microsoft KinectTM). We target application scenarios, such as warehouses or distribution and logistics companies, where it is important to promptly compute package volumes, yet high accuracy is not pivotal. Our application auto- matically detects cuboid objects using the depth camera data and computes their volume and sorting it allowing space optimization. The proposed methodology applies to a point cloud simple computer vision and image processing methods, as connected components, morphological operations and Harris corner detector, producing encouraging results, namely an accuracy in volume measurement of 8mm. Aspects that can be further improved are identified; nevertheless, the current solution is already promising turning out to be cost effective for the envisaged scenarios.

Freehand Gesture-Based 3D Manipulation Methods for Interaction with Large Displays

Gesture-based 3D interaction is a research topic with application in numerous scenarios which gained relevance with the recent advances in low-cost tracking systems. Yet, it poses many challenges due to its novelty and consequent lack of systematic development methodologies. Developing easy to use and learn gesture-based 3D interfaces is particularly difficult since the most adequate and intuitive gestures are not always obvious and there is often a variety of different gestures used to perform similar actions. This paper presents the development and evaluation of interaction methods to manipulate 3D virtual objects in a large display set-up using freehand gestures detected by a Kinect depth sensor. We describe the implementation of these methods and the user studies conducted to improve them and assess their usability as manipulation methods. Based on the results of these studies we also propose a method that overcomes the lack of roll movement detection by the Kinect and makes simpler the scaling and rotation in all degrees-of-freedom using hand gestures.

A Study on Trust in a Robotic Suitcase

This work presents a study on human-robot interaction between a prototype of a robotic suitcase – aBag – and people using it. Importantly, for an autonomous robotic suitcase to be successful as a product, people need to trust it. Therefore, a study was performed, where participants used aBag (remotely operated using the Wizard of Oz technique) for carrying their belongings. Two different conditions were created: (1) aBag follows the participant at a close range; (2) aBag follows the participant on a further distance. We expected that participants would trust more aBag when it was following them at a close range, but interestingly participants seemed to trust more when aBag was further away. Also, regardless of the conditions, the level of trust in aBag was significantly higher after the interaction compared to before, bringing positive results to the development of this kind of robotic apparatus.

Heuristic Evaluation in Information Visualization Using Three Sets of Heuristics: An Exploratory Study

Evaluation in Information Visualization is inherently complex, and it is still a challenge. Whereas it is possible to adapt evaluation methods from other fields, as Human-Computer Interaction, this adaptation may not be straightforward since visualization applications are very specific interactive systems.

FLORIS and CLORIS: Hybrid source and network localization based on ranges and video

We propose hybrid methods for localization in wireless sensor networks fusing noisy range measurements with angular information (extracted from video). Compared with conventional methods that rely on a single sensed variable, this may pave the way for improved localization accuracy and robustness. We address both the single-source and network (i.e., cooperative multiple-source) localization paradigms, solving them via optimization of a convex surrogate. The formulations for hybrid localization are unified in the sense that we propose a single nonlinear least-squares cost function, fusing both angular and range measurements. We then relax the problem to obtain an estimate of the optimal positions. This contrasts with other hybrid approaches that alternate the execution of localization algorithms for each type of measurement separately, to progressively refine the position estimates. Single-source localization uses a semidefinite relaxation to obtain a one-shot matrix solution from which the source position is derived via factorization. Network localization uses a different approach where sensor coordinates are retained as optimization variables, and the relaxed cost function is efficiently minimized using fast iterations based on Nesterovs optimal method. Further, an automated calibration procedure is developed to express range and angular information, obtained by different devices, possibly deployed at different locations, in a single consistent coordinate system. This drastically reduces the need for manual calibration that would otherwise negatively impact the practical usability of hybrid range/video localization systems. We develop and test, both in simulation and experimentally, the new hybrid localization algorithms, which not only overcome the limitations of previous fusing approaches but also compare favorably to state-of-the-art methods, outperforming them in some scenarios.

An Exploratory Study on the Predictive Capacity of Heuristic Evaluation in Visualization Applications

Heuristic evaluation is generally considered as an adequate method to perform formative usability evaluation as it helps identify potential problems from early stages of development and may provide useful results even with a relatively low investment. In particular, the method has been adapted and used to evaluate visualization applications. This paper presents an exploratory study aimed at better understanding the capacity of heuristic evaluation to predict the issues experienced by users when using a visualization application and how to assess it. The main usability potential problems pointed out in a visualization application by 20 evaluators using heuristic evaluation are compared with the problems reported for the same application by 44 users.

LocDyn: Robust Distributed Localization for Mobile Underwater Networks

How do we self-localize large teams of underwater nodes using only noisy range measurements? How do we do it in a distributed way, and incorporating dynamics into the problem? How do we reject outliers and produce trustworthy position estimates? And what if some of the vehicles can measure angular information? The stringent acoustic communication constraints and accuracy needs of our geophysical survey application demand fast and very accurate localization methods. We address dynamic localization as a MAP estimation problem where the prior encodes kinematic information, and we apply a convex relaxation method that takes advantage of previous estimates at each measurement acquisition step. The resulting LocDyn algorithm is fast: It converges at an optimal rate for first order methods. LocDyn is distributed: There is no fusion center responsible for processing acquired data and the same simple computations are performed at each node. LocDyn is accurate: Numerical experiments attest to about 30% smaller positioning error than a comparable Kalman filter. LocDyn is robust: It rejects outlier noise, while benchmarking methods succumb in terms of positioning error.