In this paper, we propose an additionsubtraction twin-gated recurrent network (ATR) to simplify neural machine translation. The recurrent units of ATR are heavily simplified to have the smallest number of weight matrices among units of all existing gated RNNs.With the simple addition and subtraction operation, we introduce a twin-gated mechanism to build input and forget gates which are highly correlated. Despite this simplification, the essential non-linearities and capability of modeling long-distance dependencies are preserved. Additionally, the proposed ATR is more transparent than LSTM/GRU due to the simplification. Forward self-attention can be easily established in ATR, which makes the proposed network interpretable. Experiments on WMT14 translation tasks demonstrate that ATR-based neural machine translation can yield competitive performance on English-German and English-French language pairs in terms of both translation quality and speed. Further experiments on NIST Chinese-English translation, natural language inference and Chinese word segmentation verify the generality and applicability of ATR on different natural language processing tasks.

The calibrated matrix of the second-order difference of the pitch delay (C-MSDPD) feature has been proven to be effective in detecting steganography based on pitch delay. In this article, a new steganalysis scheme based on multiple statistical features of pitch delay is present. Analyzing the principle of the adaptive multi-rate (AMR) codec, the pitch delay values in the same frame is divided into groups, in each of which, a pitch delay has a closer correlation with the other ones. To depict the characteristic of the pitch delay, two new types of statistical features are adopted in this article. The new features and C-MSDPD feature are together employed to train a classifier based on support vector machine (SVM). The experimental result shows that, the proposed scheme outperforms the existing one at different embedding bit rates and with different speech lengths.

Aging is an inevitable progressive decline in physiological organ function that increases the chance of disease and death. The renin–angiotensin system (RAS) is involved in the regulation of vasoconstriction, fluid homeostasis, cell growth, fibrosis, inflammation, and oxidative stress. In recent years, unprecedented advancement has been made in the RAS study, particularly with the observation that angiotensin II (Ang II), the central product of the RAS, plays a significant role in aging and chronic disease burden with aging. Binding to its receptors (Ang II type 1 receptor – AT1R in particular), Ang II acts as a mediator in the aging process by increasing free radical production and, consequently, mitochondrial dysfunction and telomere attrition. In this review, we examine the physiological function of the RAS and reactive oxygen species (ROS) sources in detail, highlighting how Ang II amplifies or drives mitochondrial dysfunction and telomere attrition underlying each hallmark of aging and contributes to the development of aging and age-linked diseases. Accordingly, the Ang II/AT1R pathway opens a new preventive and therapeutic direction for delaying aging and reducing the incidence of age-related diseases in the future.