Conditional automated driving allows drivers to perform various nondriving-related tasks (NDRTs) during commutes. This study explores how commuting direction expectations and perceived values of NDRTs impact takeover performance. We conducted a 2 × 2 mixed-design driving simulation experiment involving 64 participants, with the perceived values of NDRTs (hedonic vs. utilitarian) as a within-subject variable and commuting direction expectations (outbound vs. inbound) as a between-subject variable. The key findings include: (1) Hedonic NDRTs caused more collisions than did utilitarian NDRTs. (2) Takeovers were slower and worse during inbound commutes than during outbound commutes. (3) Goal conflicts between NDRT values and commuting direction expectations impair takeovers. Our study identifies the impact of two psychological factors, task perceived value and commuting direction expectations, on takeover behavior and highlights the impact of goal conflict. This study broadens the scope of takeover research and provides insights for designing safer in-vehicle systems for conditional automated driving.

Emotion tracking is commonly practiced for mental wellbeing.In recent years, increasing attention is turned to automatic emotion tracking, including technologies based on Facial Expression Recognition (FER).However, work on FER based tracking so far has mainly focused on accuracy improvement, and little is on how it works in everyday lives.In this paper, we developed an FER-based emotion tracking web application, EmoAction, and used it as a technology probe to explore participants' experiences and perceptions of using FER-based emotion tracking in daily life.A field study with 9 participants trying it over 14 days reveals how such an emotion tracking tool enables people to gain new awareness and new perspectives of their emotions, and identifies several potential usages of such an automatic tracking tool.We then discuss the implications for future design, including providing real-time non-judgmental feedback, prioritizing triggering reflection over seeking maximum precision, and enabling social sharing with respect for privacy.

Wrist-worn devices hold significant potential for health monitoring, particularly for older people. However, the adoption and continuous usage of these devices are heavily affected by wearing comfort. Currently, there is still a lack of validated assessment tools specifically developed to evaluate the comfort of such devices. Additionally, while pressure is critical for both proper device placement and maintaining comfort, quantitative references for suitbale pressure levels are scarce. Moreover, the impact of age difference on comfort perceptions have been largely overlooked. To address these gaps, we conducted two experiments involving 18 older and 18 younger participants. In Experiment 1, we gathered comfort evaluations from participants while wearing the devices in both home/office and exercise scenarios. In Experiment 2, we focused on pressure, measuring the pressure on participants’ wrists under three strap material conditions and collecting their corresponding comfort evaluations. The results established three dimensions of comfort: movement comfort, thermophysiological comfort, and contact comfort, with their relative importance varying by age and use scenario. Also, we identified comfort-discomfort pressure ranges for both age groups, highlighting the influence of strap material and age on comfort perceptions. These findings offer valuable design guidance for manufacturers and enhance understanding of wrist-mounted device comfort.

The emergence of Neural Radiance Fields (NeRF) has promoted the development of synthesized high-fidelity views of the intricate real world. However, it is still a very demanding task to repaint the content in NeRF. In this paper, we propose a novel framework that can take RGB images as input and alter the 3D content in neural scenes. Our work leverages existing diffusion models to guide changes in the designated 3D content. Specifically, we semantically select the target object and a pre-trained diffusion model will guide the NeRF model to generate new 3D objects, which can improve the editability, diversity, and application range of NeRF. Experiment results show that our algorithm is effective for editing 3D objects in NeRF under different text prompts, including editing appearance, shape, and more. We validate our method on both real-world datasets and synthetic-world datasets for these editing tasks. Please visit https://repaintnerf.github.io for a better view of our results.

The Social Media Prediction (SMP) challenge aims to predict the future popularity of online posts by leveraging social media data. Social media data contains multimodal information, such as text, images, time series, etc. Previous methods have proposed many feature extraction and feature construction methods to represent these multimodal information, thereby predicting the popularity of posts. Despite the success of previous methods in extracting features from social media data, these features tend to be predominantly lower-order, posing a challenge in accurately capturing the rich information contained in text and images. In this paper, we propose a more diverse feature mining method and introduce a stacking block module to capture higher-order feature information contained in text and images. "lower-order" refers to the original high-dimensional embedding representation, while "high-order" pertains to the impact on post social popularity captured by tree models from text or image. We conducted massive experiments to evaluate the effectiveness of our proposed method and found that the stacking block module significantly improved performance.

Although often ignored, adult gamers, as with children and adolescents, can suffer from problematic gaming. This study explores the use of a built-in gaming gradual intervention system (G-GIS) designed to help adult gamers achieve their desired gaming habits in an autonomous and acceptable manner. In this study, we interviewed 26 heavy adult gamers (i.e., adult gamers who played frequently and for relatively long periods of time) of an online poker game using an in-built G-GIS. Then, we triangulated the interviewees' results with their real-world behavioral data collected over eight months to explore how demographics, attitudes, and contextual factors influenced their use of the G-GIS. The results indicate that family and occupation demographics play key roles in determining adult gamers' gaming habits and their self-control under G-GIS intervention. We also revealed that adult gamers' attitudes and contextual factors can facilitate or hinder the effectiveness of the G-GIS. The findings of this study extend our understanding of heavy adult gamers and reveal how a G-GIS influences adult gamers at the individual level, which can be applied in the design of future game intervention systems, particularly for adult gamers.

The desirability of being able to search for specific persons in surveillance videos captured by different cameras has increasingly motivated interest in the problem of person re-identification, which is a critical yet under-addressed challenge in multi-camera tracking systems. The main difficulty of person re-identification arises from the variations in human appearances from different camera views. In this paper, to bridge the human appearance variations across cameras, two coupled dictionaries that relate to the gallery and probe cameras are jointly learned in the training phase from both labeled and unlabeled images. The labeled training images carry the relationship between features from different cameras, and the abundant unlabeled training images are introduced to exploit the geometry of the marginal distribution for obtaining robust sparse representation. In the testing phase, the feature of each target image from the probe camera is first encoded by the sparse representation and then recovered in the feature space spanned by the images from the gallery camera. The features of the same person from different cameras are similar following the above transformation. Experimental results on publicly available datasets demonstrate the superiority of our method.