Search

This abstract presents a scenario for human-robot action in a home setting involving an older adult and a robot. The scenario is designed to explore the envisioned modelling of memory for communication with a socially assistive robots (SAR). The scenario will enable the gathering of data on failures of speech technology and human-robot communication involving shared memory that may occur during daily activities such as a music-listening activity., Comment: In WTF Workshop Proceedings (arXiv:2401.04108) held in conjunction with the ACM conference on Conversational User Interfaces (CUI), 19 - 21/07 2023, in Eindhoven, The Netherlands

We propose methods for analysis, design, and evaluation of Meaningful Human Control (MHC) for defense technologies from the perspective of military human-machine teaming (HMT). Our approach is based on three principles. Firstly, MHC should be regarded as a core objective that guides all phases of analysis, design and evaluation. Secondly, MHC affects all parts of the socio-technical system, including humans, machines, AI, interactions, and context. Lastly, MHC should be viewed as a property that spans longer periods of time, encompassing both prior and realtime control by multiple actors. To describe macrolevel design options for achieving MHC, we propose various Team Design Patterns. Furthermore, we present a case study, where we applied some of these methods to envision HMT, involving robots and soldiers in a search and rescue task in a military context., Comment: Preprint, to be published as: van Diggelen, J., van den Bosch, K., Neerincx, M., & Steen, M. (2023). Designing for Meaningful Human Control in Military Human-Machine Teams, in Research handbook on Meaningful Human Control of Artificial Intelligence Systems. Edward Elgar Publishing

Despite their prevalence in eHealth applications for behavior change, persuasive messages tend to have small effects on behavior. Conditions or states (e.g., confidence, knowledge, motivation) and characteristics (e.g., gender, age, personality) of persuadees are two promising components for more effective algorithms for choosing persuasive messages. However, it is not yet sufficiently clear how well considering these components allows one to predict behavior after persuasive attempts, especially in the long run. Since collecting data for many algorithm components is costly and places a burden on users, a better understanding of the impact of individual components in practice is welcome. This can help to make an informed decision on which components to use. We thus conducted a longitudinal study in which a virtual coach persuaded 671 daily smokers to do preparatory activities for quitting smoking and becoming more physically active, such as envisioning one's desired future self. Based on the collected data, we designed a Reinforcement Learning (RL)-approach that considers current and future states to maximize the effort people spend on their activities. Using this RL-approach, we found, based on leave-one-out cross-validation, that considering states helps to predict both behavior and future states. User characteristics and especially involvement in the activities, on the other hand, only help to predict behavior if used in combination with states rather than alone. We see these results as supporting the use of states and involvement in persuasion algorithms. Our dataset is available online., Comment: 10 pages, 6 figures, Camera-ready version for the 22nd International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2023)

Inspired by the cognitive science theory of the explicit human memory systems, we have modeled an agent with short-term, episodic, and semantic memory systems, each of which is modeled with a knowledge graph. To evaluate this system and analyze the behavior of this agent, we designed and released our own reinforcement learning agent environment, "the Room", where an agent has to learn how to encode, store, and retrieve memories to maximize its return by answering questions. We show that our deep Q-learning based agent successfully learns whether a short-term memory should be forgotten, or rather be stored in the episodic or semantic memory systems. Our experiments indicate that an agent with human-like memory systems can outperform an agent without this memory structure in the environment.

The rapid development of Artificial Intelligence (AI) requires developers and designers of AI systems to focus on the collaboration between humans and machines. AI explanations of system behavior and reasoning are vital for effective collaboration by fostering appropriate trust, ensuring understanding, and addressing issues of fairness and bias. However, various contextual and subjective factors can influence an AI system explanation's effectiveness. This work draws inspiration from findings in cognitive psychology to understand how effective explanations can be designed. We identify four components to which explanation designers can pay special attention: perception, semantics, intent, and user & context. We illustrate the use of these four explanation components with an example of estimating food calories by combining text with visuals, probabilities with exemplars, and intent communication with both user and context in mind. We propose that the significant challenge for effective AI explanations is an additional step between explanation generation using algorithms not producing interpretable explanations and explanation communication. We believe this extra step will benefit from carefully considering the four explanation components outlined in our work, which can positively affect the explanation's effectiveness., Comment: 2022 IEEE Workshop on TRust and EXpertise in Visual Analytics (TREX)

Inspired by the cognitive science theory, we explicitly model an agent with both semantic and episodic memory systems, and show that it is better than having just one of the two memory systems. In order to show this, we have designed and released our own challenging environment, "the Room", compatible with OpenAI Gym, where an agent has to properly learn how to encode, store, and retrieve memories to maximize its rewards. The Room environment allows for a hybrid intelligence setup where machines and humans can collaborate. We show that two agents collaborating with each other results in better performance than one agent acting alone., Comment: Submitted to Human-Centered Design of Symbiotic Hybrid Intelligence 2022 (https://ii.tudelft.nl/humancenteredsymbioticHI/)

We present a discovery-based, first version, explicit model of social interaction that provides a basis for measuring the quality of interaction of a human user with a social robot. The two core elements of the social interaction model are engagement and co-regulation. Engagement emphasizes the \textit{qualitative nature} of social interaction and the fact that a user needs to be drawn into the interaction with the robot. Co-regulation emphasizes the interaction process and the fact that a user and a robot need to be acting together. We argue that the quality of social interaction with a robot can be measured in terms of how efficiently engagement and co-regulation are established and maintained during the interaction and how satisfied the user is with the interaction., Comment: Quality of Interaction in Socially Assistive Robots (QISAR) Workshop at International Conference on Social Robotics (ICSR19) - Madrid, Spain, November 26-29, 2019

A key challenge in the accurate prediction of viewers' emotional responses to video stimuli in real-world applications is accounting for person- and situation-specific variation. An important contextual influence shaping individuals' subjective experience of a video is the personal memories that it triggers in them. Prior research has found that this memory influence explains more variation in video-induced emotions than other contextual variables commonly used for personalizing predictions, such as viewers' demographics or personality. In this article, we show that (1) automatic analysis of text describing their video-triggered memories can account for variation in viewers' emotional responses, and (2) that combining such an analysis with that of a video's audiovisual content enhances the accuracy of automatic predictions. We discuss the relevance of these findings for improving on state of the art approaches to automated affective video analysis in personalized contexts.

In robotics, one of the main challenges is that the on-board Artificial Intelligence (AI) must deal with different or unexpected environments. Such AI agents may be incompetent there, while the underlying model itself may not be aware of this (e.g., deep learning models are often overly confident). This paper proposes two methods for the online assessment of the competence of the AI model, respectively for situations when nothing is known about competence beforehand, and when there is prior knowledge about competence (in semantic form). The proposed method assesses whether the current environment is known. If not, it asks a human for feedback about its competence. If it knows the environment, it assesses its competence by generalizing from earlier experience. Results on real data show the merit of competence assessment for a robot moving through various environments in which it sometimes is competent and at other times it is not competent. We discuss the role of the human in robot's self-assessment of its competence, and the challenges to acquire complementary information from the human that reinforces the assessments., Comment: Presented at the 2020 Workshop on Assessing, Explaining, and Conveying Robot Proficiency for Human-Robot Teaming

Machine Learning models become increasingly proficient in complex tasks. However, even for experts in the field, it can be difficult to understand what the model learned. This hampers trust and acceptance, and it obstructs the possibility to correct the model. There is therefore a need for transparency of machine learning models. The development of transparent classification models has received much attention, but there are few developments for achieving transparent Reinforcement Learning (RL) models. In this study we propose a method that enables a RL agent to explain its behavior in terms of the expected consequences of state transitions and outcomes. First, we define a translation of states and actions to a description that is easier to understand for human users. Second, we developed a procedure that enables the agent to obtain the consequences of a single action, as well as its entire policy. The method calculates contrasts between the consequences of a policy derived from a user query, and of the learned policy of the agent. Third, a format for generating explanations was constructed. A pilot survey study was conducted to explore preferences of users for different explanation properties. Results indicate that human users tend to favor explanations about policy rather than about single actions., Comment: XAI workshop on the IJCAI conference 2018, Stockholm, Sweden

Recent advances in interpretable Machine Learning (iML) and eXplainable AI (XAI) construct explanations based on the importance of features in classification tasks. However, in a high-dimensional feature space this approach may become unfeasible without restraining the set of important features. We propose to utilize the human tendency to ask questions like "Why this output (the fact) instead of that output (the foil)?" to reduce the number of features to those that play a main role in the asked contrast. Our proposed method utilizes locally trained one-versus-all decision trees to identify the disjoint set of rules that causes the tree to classify data points as the foil and not as the fact. In this study we illustrate this approach on three benchmark classification tasks., Comment: presented at 2018 ICML Workshop on Human Interpretability in Machine Learning (WHI 2018), Stockholm, Sweden

A reliable wireless connection between the operator and the teleoperated Unmanned Ground Vehicle (UGV) is critical in many Urban Search and Rescue (USAR) missions. Unfortunately, as was seen in e.g. the Fukushima disaster, the networks available in areas where USAR missions take place are often severely limited in range and coverage. Therefore, during mission execution, the operator needs to keep track of not only the physical parts of the mission, such as navigating through an area or searching for victims, but also the variations in network connectivity across the environment. In this paper, we propose and evaluate a new teleoperation User Interface (UI) that includes a way of estimating the Direction of Arrival (DoA) of the Radio Signal Strength (RSS) and integrating the DoA information in the interface. The evaluation shows that using the interface results in more objects found, and less aborted missions due to connectivity problems, as compared to a standard interface. The proposed interface is an extension to an existing interface centered around the video stream captured by the UGV. But instead of just showing the network signal strength in terms of percent and a set of bars, the additional information of DoA is added in terms of a color bar surrounding the video feed. With this information, the operator knows what movement directions are safe, even when moving in regions close to the connectivity threshold., Comment: Accepted for publication in the Journal of Human-Robot Interaction (JHRI)