Precision frequency measurements of terahertz (THz) waves are required to establish metrology in the THz spectral region. However, frequency measurement techniques in this region are immature. We propose a THz spectrum analyzer to measure the absolute frequency and spectral shape of continuous-wave THz waves. Based on a stable frequency comb generated in a photoconductive antenna, the absolute frequency of a sub- THz test source was determined at a precision of 2.8 x 10(-11). Furthermore, the spectral bandwidth of the THz spectrum analyzer can be extended over 1 THz, as demonstrated by measurement of a THz test source. This spectrum analyzer has the potential to become a powerful tool for THz frequency metrology.

We constructed a fiber-based, hybrid terahertz (THz) spectrometer having two working modes, asynchronous-optical-sampling THz time-domain spectroscopy (AOS-THz-TDS) and multiple-frequency-heterodyning THz comb spectroscopy (MFH-THz-CS), by use of dual fiber-laser-based frequency combs. A spectral range of 2THz and dynamic range of 100 was achieved at the single sweep measurement of 200ms in the AOS-THz-TDS mode, whereas the detailed structure of the THz frequency comb was clearly observed in the MFH-THz-CS mode. The spectrometer features compactness, robustness, flexibility, and cost effectiveness, in addition to high spectral resolution in rapid data acquisition, and has the potential to become a powerful tool for practical applications.

An automatic calibration function for a dynamic drive scheme has been developed to stabilize the contrast ratio and the number of colors for color electronic paper. We use the pixels' capacitance to detect their optical characteristics. We can accurately distinguish 4,096 colors; only a few colors were practical, previously.

We have quantified the image quality of color electronic paper (e‐paper) by subjective evaluations and clarified the requirements for improving the image quality. In the color e‐paper field, various display types have been developed with the focus on improving the display performances or manufacture costs. However, color e‐papers are still under development, and the image quality is not sufficient when compared with printed papers, photographs and conventional displays. With respect to the image quality of color e‐paper in the future, it is very significant to clarify how reflectance, contrast ratio, National Television System Committee (NTSC) ratio, and other factors affect the image quality. Therefore, we have quantified the image quality factors such as color reproduction, tone reproduction, and surface texture. From our experimental results, we could clarify the significant factors and the practical range of image quality of color e‐paper. For instance, both the reflectance of approximately 30% or higher, contrast ratio of approximately 8:1 or higher, and NTSC ratio of 20% or higher are the requirements for obtaining subjective evaluation score that is equivalent to that of newspapers, which are read very widely.

Recently, automatic speech recognition (ASR) and visual speech recognition (VSR) have been widely researched owing to development in deep learning. Most VSR research works focus only on frontal face images. However, assuming real scenes, it is obvious that a VSR system should correctly recognize spoken contents from not only frontal but also diagonal or profile faces. In this paper, we propose a novel VSR method that is applicable to faces taken at any angle. Firstly, view classification is carried out to estimate face angles. Based on the results, feature extraction is then conducted using the best combination of pre-trained feature extraction models. Next, lipreading is carried out using the features. We also developed audio-visual speech recognition (AVSR) using the VSR in addition to conventional ASR. Audio results were obtained from ASR, followed by incorporating audio and visual results in a decision fusion manner. We evaluated our methods using OuluVS2, a multi-angle audio-visual database. We then confirmed that our approach achieved the best performance among conventional VSR schemes in a phrase classification task. In addition, we found that our AVSR results are better than ASR and VSR results.

No abstract

No abstract

No abstract