Florian Alt

Chandramohan Sudar, Michael Froehlich, Florian Alt

The growing use of supervised machine learning in research and industry has increased the need for labeled datasets. Crowdsourcing has emerged as a popular method to create data labels. However, working on large batches of tasks leads to worker fatigue, negatively impacting labeling quality. To address this, we present TruEyes, a collaborative crowdsourcing system, enabling the distribution of micro-tasks to mobile app users. TruEyes allows machine learning practitioners to publish labeling tasks, mobile app developers to integrate task ads for monetization, and users to label data instead of watching advertisements. To evaluate the system, we conducted an experiment with N=296 participants. Our results show that the quality of the labeled data is comparable to traditional crowdsourcing approaches and most users prefer task ads over traditional ads. We discuss extensions to the system and address how mobile advertisement space can be used as a productive resource in the future.

Lukas Mecke, Daniel Buschek, Uwe Gruenefeld et al.

Many important decisions in our everyday lives, such as authentication via biometric models, are made by Artificial Intelligence (AI) systems. These can be in poor alignment with human expectations, and testing them on clear-cut existing data may not be enough to uncover those cases. We propose a method to find samples in the latent space of a generative model, designed to be challenging for a decision-making model with regard to matching human expectations. By presenting those samples to both the decision-making model and human raters, we can identify areas where its decisions align with human intuition and where they contradict it. We apply this method to a face recognition model and collect a dataset of 11,200 human ratings from 100 participants. We discuss findings from our dataset and how our approach can be used to explore the performance of AI models in different contexts and for different user groups.

Radiah Rivu, Ville Mäkelä, Sarah Prange et al.

We investigate the opportunities and challenges of running virtual reality (VR) studies remotely. Today, many consumers own head-mounted displays (HMDs), allowing them to participate in scientific studies from their homes using their own equipment. Researchers can benefit from this approach by being able to reach a more diverse study population and to conduct research at times when it is difficult to get people into the lab (cf. the COVID pandemic). We first conducted an online survey (N=227), assessing HMD owners' demographics, their VR setups, and their attitudes towards remote participation. We then identified different approaches to running remote studies and conducted two case studies for an in-depth understanding. We synthesize our findings into a framework for remote VR studies, discuss the strengths and weaknesses of the different approaches, and derive best practices. Our work is valuable for HCI researchers conducting VR studies outside labs.

Michael Braun, Jingyi Li, Florian Weber et al.

In this paper we present use cases for affective user interfaces (UIs) in cars and how they are perceived by potential users in China and Germany. Emotion-aware interaction is enabled by the improvement of ubiquitous sensing methods and provides potential benefits for both traffic safety and personal well-being. To promote the adoption of affective interaction at an international scale, we developed 20 mobile in-car use cases through an inter-cultural design approach and evaluated them with 65 drivers in Germany and China. Our data shows perceived benefits in specific areas of pragmatic quality as well as cultural differences, especially for socially interactive use cases. We also discuss general implications for future affective automotive UI. Our results provide a perspective on cultural peculiarities and a concrete starting point for practitioners and researchers working on emotion-aware interfaces.

Michael Braun, Florian Weber, Florian Alt

Affective technology offers exciting opportunities to improve road safety by catering to human emotions. Modern car interiors enable the contactless detection of user states, paving the way for a systematic promotion of safe driver behavior through emotion regulation. We review the current literature regarding the impact of emotions on driver behavior and analyze the state of emotion regulation approaches in the car. We summarize challenges for affective interaction in form of cultural aspects, technological hurdles and methodological considerations, as well as opportunities to improve road safety by reinstating drivers into an emotionally balanced state. The purpose of this review is to outline the community's combined knowledge for interested researchers, to provide a focussed introduction for practitioners and to identify future directions for affective interaction in the car.

Heiko Drewes, Mohamed Khamis, Florian Alt

We introduce and evaluate a novel approach for detecting smooth pursuit eye movements that increases the number of distinguishable targets and is more robust against false positives. Being natural and calibration-free, Pursuits has been gaining popularity in the past years. At the same time, current implementations show poor performance when more than eight on-screen targets are being used, thus limiting its applicability. Our approach (1) leverages the slope of a regression line, and (2) introduces a minimum signal duration that improves both the new and the traditional detection method. After introducing the approach as well as the implementation, we compare it to the traditional correlation-based Pursuits detection method. We tested the approach up to 24 targets and show that, if accepting a similar error rate, nearly twice as many targets can be distinguished compared to state of the art. For fewer targets, accuracy increases significantly. We believe our approach will enable more robust pursuit-based user interfaces, thus making it valuable for both researchers and practitioners.