Search

Art forgery is one of the important issues in art, but the general approaches using well-known image datasets are likely to fail to identify real forgeries. This is because the distributions of the well-known image datasets are very different from that of a dataset that includes the pictures created by professional counterfeiters. However, there are few datasets specialized for art forgery detection. In this research, we constructed a dataset for author identification of drawings. To create the drawings, we traced a few original sketches on top of semi-transparent papers. The traced images look more similar to the originals than typical forgery artwork images do, and it implies that it is more difficult to identify the authors of the traced images. We also suggest a new generative adversarial network model for author identification. We approach the problem by maximizing mutual information between an author and his traced artwork image. We compared our results to those obtained by the baseline convolutional neural networks. The experimental results showed that the suggested approach produced more convincing results than other approaches.

Object detection is useful for authenticating artworks, and object detection methods based on deep learning have been actively studied in recent years. However, existing object detection methods cannot perform well on artworks, due to the lack of labeled artwork datasets. To overcome the problem, we propose a novel approach for object detection on artworks. The proposed approach consists of two steps. First, we apply neural style transfer to the existing images for data augmentation. Second, we train an object detection method on the augmented dataset. We evaluate the proposed approach on Brueghel and People-Art datasets. The experimental results show that our proposed approach successfully detects objects in artworks. Moreover, our approach achieved 40.6 mAP, which is 7.8 mAP higher than the conventional approach without data augmentation. It is expected that the proposed approach is of great benefit to art historians for analyzing a large number of artworks.

Safeguarding intangible cultural heritage such as traditional painting is getting more critical, but the conventional way to preserve intangible heritage has been to simply record the performance and explanations of Living Human Treasures using video cameras. It cannot preserve the whole procedure and detailed performances of human experts. In this paper, we suggest a multisensory system that captures the core activities of Korean traditional painting while minimizing interference with their activities. The system captures the whole procedures of the activities by tracking the three-dimensional movements of the painter and the brushes, and recording information of the pigments and canvas. We captured the activities of two professional painters and analyzed the brush movements in various ways.

In this paper, we present a novel system that can construct the 3D printable facial models in a short time. To construct such 3D models, the system has a pipeline that consists of six steps — image acquisition, camera calibration, background separation, stereo matching, reconstruction of a facial surface, and adaptation of a reference model. We describe each step in the pipeline in detail. In the stereo matching step, the most important step among all, we use a hierarchical block matching which is favorable to a human face. Our system provides a unified component-based software which handles all the steps in one place in a short time.

Recent years, the GNSS has a diverse and wide application field. It provides the services that using the personal location information. But, for the some people, it is undesirable service. Therefore, these people start using the Personal Privacy Devices. A study about the impact of PPDs and effective countermeasure to PPDs are required. In this paper, we analyze the effect on the anti-jamming technique by the PPD jammer. And we show the effect of frequency swept jammer is removed by applying reduced-rank approach. We also analyzed the performance of reduced-rank approach.

This paper proposes a wireless LAN-based digital photo frame system that displays photographs with non-photorealistic rendering (NPR) effects. The system accepts a photograph as an input and generates an artistic image using NPR techniques. Then, it displays the final images on a digital photo frame via wireless LAN. By using this system, users can easily view different artistic images by inputting a photograph.

It is commonly agreed that architecture design is essential for development of complicated software and that depending on viewpoints software architecture can be captured in various views. Among them, the logical view is the earliest and foremost view that software developers should consider because it shapes the later phase of architecture design and the subsequent development. In this paper, we present an architecture description language for logical architecture and its associated design method. The main characteristic of the language and the design method are that they support recursive top-down design so that logical architectures for software systems of any scale and complexity can be designed. We apply them to a realistic system and show their efficacy.

In this paper, a systematic approach of developing detail OO designs from component and connector architectures (CCAs) is proposed. In this approach, an intermediate model between the architecture model and the detail design model specified with class diagrams or sequence diagrams is introduced to narrow the wide gap between the two abstraction levels. Once a CCA is designed, candidate classes and their relationships are identified per each architectural element. In order to show the efficacy of this approach, we apply it to an industry software development project and verify that quality attributes supported by the CCA are equally maintained by the detail design.