Explainable AI (XAI) is widely viewed as a sine qua non for ever-expanding AI research. A better understanding of the needs of XAI users, as well as human-centered evaluations of explainable models are both a necessity and a challenge. In this paper, we explore how human-computer interaction (HCI) and AI researchers conduct user studies in XAI applications based on a systematic literature review. After identifying and thoroughly analyzing 97 core papers with human-based XAI evaluations over the past five years, we categorize them along the measured characteristics of explanatory methods, namely trust, understanding, usability , and human-AI collaboration performance . Our research shows that XAI is spreading more rapidly in certain application domains, such as recommender systems than in others, but that user evaluations are still rather sparse and incorporate hardly any insights from cognitive or social sciences. Based on a comprehensive discussion of best practices, i.e., common models, design choices, and measures in user studies, we propose practical guidelines on designing and conducting user studies for XAI researchers and practitioners. Lastly, this survey also highlights several open research directions, particularly linking psychological science and human-centered XAI.

Student engagement is a key component of learning and teaching, resulting in a plethora of automated methods to measure it. Whereas most of the literature explores student engagement analysis using computer-based learning often in the lab, we focus on using classroom instruction in authentic learning environments. We collected audiovisual recordings of secondary school classes over a one and a half month period, acquired continuous engagement labeling per student (N=15) in repeated sessions, and explored computer vision methods to classify engagement from facial videos. We learned deep embeddings for attentional and affective features by training Attention-Net for head pose estimation and Affect-Net for facial expression recognition using previously-collected large-scale datasets. We used these representations to train engagement classifiers on our data, in individual and multiple channel settings, considering temporal dependencies. The best performing engagement classifiers achieved student-independent AUCs of .620 and .720 for grades 8 and 12, respectively, with attention-based features outperforming affective features. Score-level fusion either improved the engagement classifiers or was on par with the best performing modality. We also investigated the effect of personalization and found that only 60 seconds of person-specific data, selected by margin uncertainty of the base classifier, yielded an average AUC improvement of .084.

Large language models represent a significant advancement in the field of AI. The underlying technology is key to further innovations and, despite critical views and even bans within communities and regions, large language models are here to stay. This position paper presents the potential benefits and challenges of educational applications of large language models, from student and teacher perspectives. We briefly discuss the current state of large language models and their applications. We then highlight how these models can be used to create educational content, improve student engagement and interaction, and personalize learning experiences. With regard to challenges, we argue that large language models in education require teachers and learners to develop sets of competencies and literacies necessary to both understand the technology as well as their limitations and unexpected brittleness of such systems. In addition, a clear strategy within educational systems and a clear pedagogical approach with a strong focus on critical thinking and strategies for fact checking are required to integrate and take full advantage of large language models in learning settings and teaching curricula. Other challenges such as the potential bias in the output, the need for continuous human oversight, and the potential for misuse are not unique to the application of AI in education. But we believe that, if handled sensibly, these challenges can offer insights and opportunities in education scenarios to acquaint students early on with potential societal biases, criticalities, and risks of AI applications. We conclude with recommendations for how to address these challenges and ensure that such models are used in a responsible and ethical manner in education.

Post-chiasmal visual pathway lesions and glaucomatous optic neuropathy cause binocular visual field defects (VFDs) that may critically interfere with quality of life and driving licensure. The aims of this study were (i) to assess the on-road driving performance of patients suffering from binocular visual field loss using a dual-brake vehicle, and (ii) to investigate the related compensatory mechanisms. A driving instructor, blinded to the participants' diagnosis, rated the driving performance (passed/failed) of ten patients with homonymous visual field defects (HP), including four patients with right (HR) and six patients with left homonymous visual field defects (HL), ten glaucoma patients (GP), and twenty age and gender-related ophthalmologically healthy control subjects (C) during a 40-minute driving task on a pre-specified public on-road parcours. In order to investigate the subjects' visual exploration ability, eye movements were recorded by means of a mobile eye tracker. Two additional cameras were used to monitor the driving scene and record head and shoulder movements. Thus this study is novel as a quantitative assessment of eye movements and an additional evaluation of head and shoulder was performed. Six out of ten HP and four out of ten GP were rated as fit to drive by the driving instructor, despite their binocular visual field loss. Three out of 20 control subjects failed the on-road assessment. The extent of the visual field defect was of minor importance with regard to the driving performance. The site of the homonymous visual field defect (HVFD) critically interfered with the driving ability: all failed HP subjects suffered from left homonymous visual field loss (HL) due to right hemispheric lesions. Patients who failed the driving assessment had mainly difficulties with lane keeping and gap judgment ability. Patients who passed the test displayed different exploration patterns than those who failed. Patients who passed focused longer on the central area of the visual field than patients who failed the test. In addition, patients who passed the test performed more glances towards the area of their visual field defect. In conclusion, our findings support the hypothesis that the extent of visual field per se cannot predict driving fitness, because some patients with HVFDs and advanced glaucoma can compensate for their deficit by effective visual scanning. Head movements appeared to be superior to eye and shoulder movements in predicting the outcome of the driving test under the present study scenario.

New generation head-mounted displays, such as VR and AR glasses, are coming into the market with already integrated eye tracking and are expected to enable novel ways of human-computer interaction in numerous applications. However, since eye movement properties contain biometric information, privacy concerns have to be handled properly. Privacy-preservation techniques such as differential privacy mechanisms have recently been applied to eye movement data obtained from such displays. Standard differential privacy mechanisms; however, are vulnerable due to temporal correlations between the eye movement observations. In this work, we propose a novel transform-coding based differential privacy mechanism to further adapt it to the statistics of eye movement feature data and compare various low-complexity methods. We extend the Fourier perturbation algorithm, which is a differential privacy mechanism, and correct a scaling mistake in its proof. Furthermore, we illustrate significant reductions in sample correlations in addition to query sensitivities, which provide the best utility-privacy trade-off in the eye tracking literature. Our results provide significantly high privacy without any essential loss in classification accuracies while hiding personal identifiers.