Search

Polyploid is considered an advantage that has evolved to be more environmentally adaptable than its diploid. To understand if doubled chromosome of diploid rice can improve drought tolerance, we evaluated the diploid (2X) and autotetraploid (4X) plants of three indica and three japonica varieties. Drought stress in the plastic bucket of four-leaf stage revealed that the drought tolerance of 4X plants was lower than that of its diploid donor plants. The assay of photosynthetic rate of all varieties showed that all 4X varieties had lower rates than their diploid donors. The capacity for reactive oxygen species production and scavenging varied among different 2X and 4X varieties. Further, transcriptomic analysis of 2X and 4X plants of four varieties under normal and drought condition showed that the wide variation of gene expression was caused by difference of varieties, not by chromosome ploidy. However, weighted gene co-expression network analysis (WGCNA) revealed that the severe interference of photosynthesis-related genes in tetraploid plants under drought stress is the primary reason for the decrease of drought tolerance in autotetraploid lines. Consistently, new transcripts analysis in autotetraploid revealed that the gene transcription related with mitochondrion and plastid of cell component was influenced most significantly. The results indicated that chromosome doubling of diploid rice weakened their drought tolerance, primarily due to disorder of photosynthesis-related genes in tetraploid plants under drought stress. Maintain tetraploid drought tolerance through chromosome doubling breeding in rice needs to start with the selection of parental varieties and more efforts. [ABSTRACT FROM AUTHOR]

With the introduction of emerging technologies such as cloud computing, the railway communication network has the characteristics of complex structure and blurred boundaries, which leads to a series of security threats including information leakage and malicious access. Specifically, the third-party cloud services are difficult to be supervised, and network traffic is untrustworthy. To ensure system security, we propose a zero-trust security model in this paper. Then, we introduce blockchain and Merkle tree to build a distributed identity storage scheme for guaranteeing reliable, confidential and efficient data updates, and improving authentication efficiency. Furthermore, the proxy was introduced for two-way authentication with cloud servers, so that internal and external threats could be counteracted. Moreover, reputation assessment mechanism has been adopted to reduce the possibility of nodes accessing malicious cloud services. Performance analysis demonstrated that the proposed security model is able to enhance the security, efficiency and stability of the system, and consequently can guarantee the safety and reliability of railway transportation. [ABSTRACT FROM AUTHOR]

Text-independent speaker recognition is widely used in identity recognition that has a wide spectrum of applications, such as criminal investigation, payment certification, and interest-based customer services. In order to improve the recognition ability of log filter bank feature vectors, a method of text-independent speaker recognition based on deep residual networks model was proposed in this paper. The deep residual network was composed of a residual network (ResNet) and a convolutional attention statistics pooling (CASP) layer. The CASP layer could aggregate frame-level features from the ResNet into an utterance-level features. Extracting speech features for each speaker using deep residual networks was a promising direction to explore, and a straightforward solution was to train the discriminative feature extraction network by using a margin-based loss function. However, a margin-based loss function often has certain limitations, such as the margins between different categories were set to be the same and fixed. Thus, we used an adaptive curriculum learning loss (ACLL) to address the problem and introduce two different margin-based losses for this problem, i.e., AM-Softmax and AAM-Softmax. The proposed method was applied to a large-scale VoxCeleb2 dataset for extensive text-independent speaker recognition experiments, and average equal error rate (EER) could achieve 1.76% on VoxCeleb1 test dataset, 1.91% on VoxCeleb1-E test dataset, and 3.24% on VoxCeleb1-H test dataset. Compared with related speaker recognition methods, EER was improved by 1.11% on VoxCeleb1 test dataset, 1.04% on VoxCeleb1-E test dataset, and 1.69% on VoxCeleb1-H test dataset. [ABSTRACT FROM AUTHOR]