Recently, it is expected that the systems for automatically adding the suitable music to a target image will be developed. Toward realizing the systems, we propose a method to estimate the degree of feeling that natural objects appear in the target (photographic) images. The method uses neural network-based frameworks, and the network is trained by approximately 900 images. Evaluations were conducted by a task which classifies the impression of input images as “natural” or “artificial”. The best classification rate of 76.5% is observed when using “shape information (Bag-of-Keypoints)”, “the number of object candidates”, and “the number and area of detected face regions” as input features.
In this paper, we propose a method to improve reconstruction precision in Background Subtraction from Compressive Measurements (BSCM) which is one of the framework of background extraction methods for fixed viewpoint video. The conventional BSCM method realized highly accurate reconstruction and background extraction using spatio-temporal correlation in video tensor. However, the foreground component in reconstructed video becomes unclear and the precision deteriorates. We improve the reconstruction precision by employing multiple updates of variables in optimization.
In this paper, we present a method of compressive sensing and recovery of binary images based on Discrete Cosine Transform (DCT). Conventional method based on Hadamard transform well recovers a given binary image, but image accuracy gets worse in the case of small number of input information or complex input images. We develop a novel method by replacing the Hadamard transformation matrix by DCT matrix, and evaluate its image recovery accuracy in comparison with the conventional method.
Recently, NII and Stanford researchers independently showed that the academic abilities of school students in Japan and US have declined significantly due to their acceptance of mobile Internet services and fragmented information. We describe the importance of providing children electronic materials appropriate for their age from the perspective of development; logical information and reality of visual stimuli. Our solution involves strengthening structural- and logical-mental functions through the bi-modal enhancement of note-taking and review processes.