With the recent gradual reduction in negative perceptions of virtual influencers (VIs), understanding how consumers perceive VIs’ anthropomorphism and its impact on forming consumer relationships is critical for increasing marketing effectiveness. To gain a better understanding of how consumers form relationships with VIs, this study examined VI characteristics as well as consumer personalities. The study’s data were collected from a sample of 266 social media users. The results show that VI’s anthropomorphism positively affected the parasocial relationship (PSR). This association was reinforced by homophily, whereas the neede for uniqueness dampened it. PSR positively influenced trust, intention to imitate, and word-of-mouth (WOM). Moreover, trust positively affected imitation intention and WOM. This study aimed to broaden theoretical understanding, specifically on consumers’ responses to VIs, which is an emerging research area, while providing practical recommendations for companies incorporating VIs into their strategies.

As Extended Reality (XR) environments increasingly involve large amounts of data and content, designing effective exploration techniques has become critical. Depth-based object exploration is a common but underexplored task in XR environments, especially in settings without physical devices. Prior studies have largely focused on horizontal or planar interactions, leaving depthoriented exploration relatively overlooked. To bridge this gap, we propose three linearly aligned layer transition techniques (Continuous Push, Push&Return, and Tilt&Return) specifically designed to support efficient, precise, and continuous object exploration along the depth axis within depth-based UIs. In a user study with 30 participants, we compared their performance, usability, and user preference across two different layer configurations (8-layer vs. 16-layer). The results highlight that Continuous Push enables faster exploration with lower effort, while Push&Return provides the highest accuracy and is most preferred by users. Based on these findings, we discuss design implications for depth-based interaction techniques in controller-free XR environments.

Abstract Playing a massively multiplayer online role-playing game (MMORPG) is no longer just a personal hobby, and the genre has evolved into a forum that encourages continual interaction among its users. To determine the motivation behind the user’s participation of MMORPG, we suggest social comparison and self-expansion theories. Specifically, this study proposes a research model that comprehensively identifies the effects of psychological motives (social comparison motive and belongingness) and immersive gaming behaviors (social presence and flow experience) on behavioral factors (continuous play intention and loyalty). We analyzed 274 valid survey questionnaire responses. Our findings contribute to a growing body of literature that identifies the motives of users to play MMORPGs by applying social comparison and self-expansion theories to understand the factors that influence people to play MMORPGs. Furthermore, the study’s results have practical implications for game developers, suggesting that they should develop content strategically to attract users.

It is well known that most of the conventional video question answering (VideoQA) datasets consist of easy questions requiring simple reasoning processes. However, long videos inevitably contain complex and compositional semantic structures along with the spatio-temporal axis, which requires a model to understand the compositional structures inherent in the videos. In this paper, we suggest a new compositional VideoQA method based on transformer architecture with a deformable attention mechanism to address the complex VideoQA tasks. The deformable attentions are introduced to sample a subset of informative visual features from the dense visual feature map to cover a temporally long range of frames efficiently. Furthermore, the dependency structure within the complex question sentences is also combined with the language embeddings to readily understand the relations among question words. Extensive experiments and ablation studies show that the suggested dense but efficient model outperforms other baselines.

This paper proposes a novel scheme to detect and localize the spy cameras based on AI algorithm based raw traffic analytics, named AI-aided Hidden Camera Locator (AHCL). In AHCL, the video streaming data are filtered via the SVM (support vector machine) algorithm to quickly monitor whole raw network traffic from a router to the networks first. Then, gathered traffic data are denoised by the Denoising Autoencoder (DAE) technique to improve the data quality of classification for localization, where a camera transmits video streaming. Based on the proof-of-concept implementation, the proposed scheme can achieve 99.5% positioning accuracy of camera detection with the Ensemble Neural Networks (NNs).

Deep neural networks often make decisions based on the spurious correlations inherent in the dataset, failing to generalize in an unbiased data distribution. Although previous approaches pre-define the type of dataset bias to prevent the network from learning it, recognizing the bias type in the real dataset is often prohibitive. This paper proposes a novel bias-tailored augmentation-based approach, BiaSwap, for learning debiased representation without requiring supervision on the bias type. Assuming that the bias corresponds to the easy-to-learn attributes, we sort the training images based on how much a biased classifier can exploits them as shortcut and divide them into bias-guiding and bias-contrary samples in an unsupervised manner. Afterwards, we integrate the style-transferring module of the image translation model with the class activation maps of such biased classifier, which enables to primarily transfer the bias attributes learned by the classifier. Therefore, given the pair of bias-guiding and bias-contrary, BiaSwap generates the bias-swapped image which contains the bias attributes from the bias-contrary images, while preserving bias-irrelevant ones in the bias-guiding images. Given such augmented images, BiaSwap demonstrates the superiority in debiasing against the existing baselines over both synthetic and real-world datasets. Even without careful supervision on the bias, BiaSwap achieves a remarkable performance on both unbiased and bias-guiding samples, implying the improved generalization capability of the model.

Image classification models tend to make decisions based on peripheral attributes of data items that have strong correlation with a target variable (i.e., dataset bias). These biased models suffer from the poor generalization capability when evaluated on unbiased datasets. Existing approaches for debiasing often identify and emphasize those samples with no such correlation (i.e., bias-conflicting) without defining the bias type in advance. However, such bias-conflicting samples are significantly scarce in biased datasets, limiting the debiasing capability of these approaches. This paper first presents an empirical analysis revealing that training with "diverse" bias-conflicting samples beyond a given training set is crucial for debiasing as well as the generalization capability. Based on this observation, we propose a novel feature-level data augmentation technique in order to synthesize diverse bias-conflicting samples. To this end, our method learns the disentangled representation of (1) the intrinsic attributes (i.e., those inherently defining a certain class) and (2) bias attributes (i.e., peripheral attributes causing the bias), from a large number of bias-aligned samples, the bias attributes of which have strong correlation with the target variable. Using the disentangled representation, we synthesize bias-conflicting samples that contain the diverse intrinsic attributes of bias-aligned samples by swapping their latent features. By utilizing these diversified bias-conflicting features during the training, our approach achieves superior classification accuracy and debiasing results against the existing baselines on both synthetic and real-world datasets.