Adversarially Learned Total Variability Embedding for Speaker Recognition with Random Digit Strings.

Over the recent years, various research has been conducted to investigate methods for verifying users with a short randomized pass-phrase due to the increasing demand for voice-based authentication systems. In this paper, we propose a novel technique for extracting an i-vector-like feature based on an adversarially learned inference (ALI) model which summarizes the variability within the Gaussian mixture model (GMM) distribution through a nonlinear process. Analogous to the previously proposed variational autoencoder (VAE)-based feature extractor, the proposed ALI-based model is trained to generate the GMM supervector according to the maximum likelihood criterion given the Baum-Welch statistics of the input utterance. However, to prevent the potential loss of information caused by the Kullback-Leibler divergence (KL divergence) regularization adopted in the VAE-based model training, the newly proposed ALI-based feature extractor exploits a joint discriminator to ensure that the generated latent variable and the GMM supervector are more realistic. The proposed framework is compared with the conventional i-vector and VAE-based methods using the TIDIGITS dataset. Experimental results show that the proposed method can represent the uncertainty caused by the short duration better than the VAE-based method. Furthermore, the proposed approach has shown great performance when applied in association with the standard i-vector framework.

Estimation of Whole Plant Photosynthetic Rate of Irwin Mango under Artificial and Natural Lights Using a Three-Dimensional Plant Model and Ray-Tracing.

Photosynthesis is an important physiological response for determination of CO2 fertilization in greenhouses and estimation of crop growth. In order to estimate the whole plant photosynthetic rate, it is necessary to investigate how light interception by crops changes with environmental and morphological factors. The objectives of this study were to analyze plant light interception using a three-dimensional (3D) plant model and ray-tracing, determine the spatial distribution of the photosynthetic rate, and estimate the whole plant photosynthetic rate of Irwin mango (Mangifera indica L. cv. Irwin) grown in greenhouses. In the case of mangoes, it is difficult to measure actual light interception at the canopy level due to their vase shape. A two-year-old Irwin mango tree was used to measure the whole plant photosynthetic rate. Light interception and whole plant photosynthetic rate were measured under artificial and natural light conditions using a closed chamber (1 × 1 × 2 m). A 3D plant model was constructed and ray-tracing simulation was conducted for calculating the photosynthetic rate with a two-variable leaf photosynthetic rate model of the plant. Under artificial light, the estimated photosynthetic rate increased from 2.0 to 2.9 µmolCO2·m-2·s-1 with increasing CO2 concentration. On the other hand, under natural light, the photosynthetic rate increased from 0.2 µmolCO2·m-2·s-1 at 06:00 to a maximum of 7.3 µmolCO2·m-2·s-1 at 09:00, then gradually decreased to -1.0 µmolCO2·m-2·s-1 at 18:00. In validation, simulation results showed good agreement with measured results with R² = 0.79 and RMSE = 0.263. The results suggest that this method could accurately estimate the whole plant photosynthetic rate and be useful for pruning and adequate CO2 fertilization.

Quantification of Spectral Perception of Plants with Light Absorption of Photoreceptors.

Although plant responses to artificial lighting spectra often produce abnormal morphogenesis and reduced productivity, no quantification method to determine how plants perceive and respond to light has been available. Our objective in this study was to test whether a plant's spectral perception can be quantified using the light absorption of its major photoreceptors, phytochrome, cryptochrome, and phototropin. We developed an artificial solar lamp and three different light sources, based on a high-pressure sodium lamp, a fluorescent lamp, and red and blue light-emitting diodes, whose absorption by photoreceptors was equal to that of the standard solar spectrum. Cucumber plants grown under the artificial solar and developed light sources showed normal photomorphogenesis and were indistinguishable from each other. Plants grown under unmodified commercial light sources had abnormal photomorphogenesis that made them short and small. The photosynthetic rate was higher under the unmodified light sources; however, dry masses were highest under the artificial solar and modified light sources, indicating that the cucumber plants are optimized to the solar spectrum. Our results clearly demonstrate that the spectral perceptions of plants can be quantified using the light absorption of their photoreceptors, not visual color or spectra. We expect that our findings will contribute to a better understanding of plant perceptions of and responses to light quality, and improve the productivity of plants cultivated under artificial light.