The use of Personal Voice Assistants (PVAs) 1 such as Alexa and the Google Assistant is rising steadily, but there is a lack of research investigating common issues and requests of PVAs in the context of smart home control. We address this research question with an online survey (n = 65), using a qualitative evaluation of users' problems and improvement requests. Our analysis leads to a partly hierarchic clustering of issues & recommendations for interaction capabilities of PVAs into seven basic categories, allowing us in turn to derive implications and to condense them into design guidelines for future Human-Agent Interaction (HAI) with PVAs. Specifically, we formulate and elaborate the concepts Authentication & Authorization, Activity-Based Interaction, Situated Dialogue, and Explainability & Transparency as key topics for making progress towards smooth interaction with smart homes.

In this work, the creation of a dataset labeled in a pixel-wise manner for the uncommon domain of stain detection on patterned laundry is described. The unique properties of images in this dataset—stains are small and sometimes occur in large amounts—led to the creation of noisy labels. Indeed, the training of a fully convolutional neural network for salient object detection with this dataset revealed that the model predicts stains missed by human labelers. Thus, the reduction in label noise by adding overlooked regions with the help of the model’s predictions is examined in two different experiments. In the model-assisted labeling experiment, a simulation is ran where a human selects correct regions from the predictions. In the self-training experiment, regions of high certainty are automatically selected from the predictions. Re-training the model with the revised labels shows that model-assisted labeling leads to an average improvement in performance by 8.52%. In contrast, with self-training, the performance increase is generally lower (2.58% on average) and a decrease is even possible since regions of high certainty are often false positives.

No abstract

No abstract