The objectives of this study are to explore the information source, assessment, and preferred styles of 3D virtual influencers(VI), to investigate the expected impact of advertisements with 3D VIs on brands, and to explore ways of expanding the use of 3D VIs. In-depth interviews with 40 males and females in their 20s and 30s were conducted and qualitative data were analyzed. The study results are summarized as follows. First, the information source of the 3D VI was SNS, acquaintances, and broadcasting. Second, 3D VIs were considered positively due to their attractive appearance, wide utilization, innovative use, freshness, separation from private identity, and time and cost savings, while considered negatively due to their unrealistic appearance and antipathy against replacing a person’s role. Third, the preferred appearance styles of the 3D VI differed according to the level of virtuality although the majority of interviewees preferred similar looks to real people with low virtuality. Fourth, diverse image qualities such as innovative, differentiated, trendy, high-value, professional, and future-oriented were considered as transferred to the brand advertised by 3D VIs. Fifth, advertisements with 3D VIs may help build positive perceptions of advertised brands that may lead to purchase behaviors for some consumers. Lastly, to expand the use of 3D VIs, the specific advantages of virtual models should be maximized with consideration of how to implement a variety of body types and images of models. Findings present an important foundation to generate strategies to better apply 3D VIs to the fashion market.

In this paper, we introduce the methodology that utilizes deep learning-based front-end to enhance underwater feature matching. Both optical camera and sonar are widely applicable sensors in underwater research, however, each sensor has its own weaknesses, such as light condition and turbidity for the optic camera, and noise for sonar. To overcome the problems, we proposed the opti-acoustic transformation method. Since feature detection in sonar image is challenging, we converted the sonar image to an optic style image. Maintaining the main contents in the sonar image, CNN-based style transfer method changed the style of the image that facilitates feature detection. Finally, we verified our result using cosine similarity comparison and feature matching against the original optic image.

In this paper, we present auto-annotation tool and synthetic dataset using 3D CAD model for deep learning based object detection. To be used as training data for deep learning methods, class, segmentation, bounding-box, contour, and pose annotations of the object are needed. We propose an automated annotation tool and synthetic image generation. Our resulting synthetic dataset reflects occlusion between objects and applicable for both underwater and in-air environments. To verify our synthetic dataset, we use MASK R-CNN as a state-of-the-art method among object detection model using deep learning. For experiment, we make the experimental environment reflecting the actual underwater environment. We show that object detection model trained via our dataset show significantly accurate results and robustness for the underwater environment. Lastly, we verify that our synthetic dataset is suitable for deep learning model for the underwater environments.