Presentation attack detection based on two-stream vision transformers with self-attention fusion

Aiming at the performance degradation of the existing presentation attack detection methods due to the illumination variation, a two-stream vision transformers framework (TSViT) based on transfer learning in two complementary spaces is proposed in this paper. The face images of RGB color space and multi-scale retinex with color restoration (MSRCR) space are fed to TSViT to learn the distinguishing features of presentation attack detection. To effectively fuse features from two sources (RGB color space images and MSRCR images), a feature fusion method based on self-attention is built, which can effectively capture the complementarity of two features. Experiments and analysis on Oulu-NPU, CASIA-MFSD, and Replay-Attack databases show that it outperforms most existing methods in intra-database testing and achieves good generalization performance in cross-database testing.     

Directional separation of hydrogen-containing gas mixture by hydrate-membrane coupling method

Recovery of hydrogen (H₂) from H₂-containing gas mixtures has great significance for energy conservation, cost reduction and benefit increase. However, the common separation methods have the ubiquitous problem due to phase equilibrium principle and results in the conflict between H₂ concentration and H₂ recovery rate in the product gas. Consequently, an innovative conception of hydrate-membrane coupling approach is proposed in this work. In the separation process, hydration and membrane permeation two separation driving forces coexist to achieve the aim of strengthening mass transfer kinetics. H₂ and non-H₂ components (hydrocarbons) are synchronously and directionally selected by membrane and hydrate to improve different phase compositions. Therefore, the gas in feed side could keep relatively high two separation driving forces (H₂ fugacity and hydrocarbons fugacity). The results show that the coupling method could synchronously increase both the concentration and the recovery rate of H₂ in the product gas. At the same time, the volume and concentration of the hydrocarbons in hydrate both increases effectively. It indicates that hydrate and membrane separation methods support each other in the separation process. The hydrate-membrane coupling method fundamentally solves the issue of the decreasing driving force resulting from single separation method and phase equilibrium relationship.     

Breeding schemes with optimum-contribution selection or truncation selection for beef cattle destined for use on dairy females

We tested the hypothesis that the size of a beef cattle population destined for use on dairy females is smaller under optimum-contribution selection (OCS) than under truncation selection (TRS) at the same genetic gain (ΔG) and the same rate of inbreeding (ΔF). We used stochastic simulation to estimate true ΔG realized at a 0.005 ΔF in breeding schemes with OCS or TRS. The schemes for the beef cattle population also differed in the number of purebred offspring per dam and the total number of purebred offspring per generation. Dams of the next generation were exclusively selected among the one-year-old heifers. All dams were donors for embryo transfer and produced a maximum of 5 or 10 offspring. The total number of purebred offspring per generation was: 400, 800, 1,600 or 4,000 calves, and it was used as a measure of population size. Rate of inbreeding was predicted and controlled using pedigree relationships. Each OCS (TRS) scheme was simulated for 10 discrete generations and replicated 100 (200) times. The OCS scheme and the TRS scheme with a maximum of 10 offspring per dam required approximately 783 and 1,257 purebred offspring per generation to realize a true ΔG of €14 and a ΔF of 0.005 per generation. Schemes with a maximum of 5 offspring per dam required more purebred offspring per generation to realize a similar true ΔG and a similar ΔF. Our results show that OCS and multiple ovulation and embryo transfer act on selection intensity through different mechanisms to achieve fewer selection candidates and fewer selected sires and dams than under TRS at the same ΔG and a fixed ΔF. Therefore, we advocate the use of a breeding scheme with OCS and multiple ovulation and embryo transfer for beef cattle destined for use on dairy females because it is favorable both from an economic perspective and a carbon footprint perspective.     

A Smart Deep Convolutional Neural Network for Real-Time Surface Inspection

A proper detection and classification of defects in steel sheets in real time have become a requirement for manufacturing these products, largely used in many industrial sectors. However, computers used in the production line of small to medium size companies, in general, lack performance to attend real-time inspection with high processing demands. In this paper, a smart deep convolutional neural network for using in real-time surface inspection of steel rolling sheets is proposed. The architecture is based on the state-of-the-art SqueezeNet approach, which was originally developed for usage with autonomous vehicles. The main features of the proposed model are: small size and low computational burden. The model is 10 to 20 times smaller when compared to other networks designed for the same task, and more than 700 times smaller than general networks. Also, the number of floating-point operations for a prediction is about 50 times lower than the ones used for similar tasks. Despite its small size, the proposed model achieved near-perfect accuracy on a public dataset of 1800 images of six types of steel rolling defects.     

安卓平台上的医用图像信息采集系统的开发设计

为了探讨在安卓平台上构建医用图像采集系统的开发个案,分析通过以智能手机、平板电脑为核心安卓设备通过拍照获得化验单数据后进行文本识别并提交智慧医疗系统的解决方案。本文首先通过二值化算法形成低阈值图像数据,使用卷积神经元网络算法对文本进行逐一识别,使用K-means算法对识别后的单字文本进行字段记录值的整合并形成元数据库服务于其他智慧医疗系统模块。在使用9000组数据对神经元网络进行前期训练的前提下,该系统的识别准确率达到了99.5%以上。本系统具有一定的可行性,对未来智慧医疗的系统开发有实践意义。     

The influence of textese on adolescents’ perceptions of text message writers

Text messages are characterised by a casual language style, ‘textese’ (e.g., c u on thurs). This study investigated adolescents’ perceptions of the use of different levels of textese in digital messages which varied in their intended recipient (friend, teacher). Grade 8 students in Australia (N = 90, aged 13–14 years) each read six text messages purportedly written by a fellow student, and rated the extent to which they agreed (on a 7-point Likert scale) that the message writer was intelligent, paid attention to detail, used an appropriate writing style, and was friendly. Overall, participants rated message senders most favourably on all measures when they used no textese, and least favourably when they used high levels of textese. This pattern was even stronger when messages were addressed to teachers rather than same-aged friends. The findings suggest that adolescents are sensitive to both writing style and recipient when considering digital messages.     

Sequential alignment attention model for scene text recognition

Scene text recognition has been a hot research topic in computer vision due to its various applications. The state-of-the-art solutions usually depend on the attention-based encoder-decoder framework that learns the mapping between input images and output sequences in a purely data-driven way. Unfortunately, there often exists severe misalignment between feature areas and text labels in real-world scenarios. To address this problem, this paper proposes a sequential alignment attention model to enhance the alignment between input images and output character sequences. In this model, an attention gated recurrent unit (AGRU) is first devised to distinguish the text and background regions, and further extract the localized features focusing on sequential text regions. Furthermore, CTC guided decoding strategy is integrated into the popular attention-based decoder, which not only helps to boost the convergence of the training but also enhances the well-aligned sequence recognition. Extensive experiments on various benchmarks, including the IIIT5k, SVT, and ICDAR datasets, show that our method substantially outperforms the state-of-the-art methods.     

Sofosbuvir Selects for Drug-Resistant Amino Acid Variants in the Zika Virus RNA-Dependent RNA-Polymerase Complex In Vitro

The nucleotide analog sofosbuvir, licensed for the treatment of hepatitis C, recently revealed activity against the Zika virus (ZIKV) in vitro and in animal models. However, the ZIKV genetic barrier to sofosbuvir has not yet been characterized. In this study, in vitro selection experiments were performed in infected human hepatoma cell lines. Increasing drug pressure significantly delayed viral breakthrough (p = 0.029). A double mutant in the NS5 gene (V360L/V607I) emerged in 3 independent experiments at 40–80 µM sofosbuvir resulting in a 3.9 ± 0.9-fold half- maximal inhibitory concentration (IC₅₀) shift with respect to the wild type (WT) virus. A triple mutant (C269Y/V360L/V607I), detected in one experiment at 80 µM, conferred a 6.8-fold IC₅₀ shift with respect to the WT. Molecular dynamics simulations confirmed that the double mutant V360L/V607I impacts the binding mode of sofosbuvir, supporting its role in sofosbuvir resistance. Due to the distance from the catalytic site and to the lack of reliable structural data, the contribution of C269Y was not investigated in silico. By a combination of sequence analysis, phenotypic susceptibility testing, and molecular modeling, we characterized a double ZIKV NS5 mutant with decreased sofosbuvir susceptibility. These data add important information to the profile of sofosbuvir as a possible lead for anti-ZIKV drug development.