A blockchain-based audit approach for encrypted data in federated learning

The development of data-driven artificial intelligence technology has given birth to a variety of big data applications. Data has become an essential factor to improve these applications. Federated learning, a privacy-preserving machine learning method, is proposed to leverage data from different data owners. It is typically used in conjunction with cryptographic methods, in which data owners train the global model by sharing encrypted model updates. However, data encryption makes it difficult to identify the quality of these model updates. Malicious data owners may launch attacks such as data poisoning and free-riding. To defend against such attacks, it is necessary to find an approach to audit encrypted model updates. In this paper, we propose a blockchain-based audit approach for encrypted gradients. It uses a behavior chain to record the encrypted gradients from data owners, and an audit chain to evaluate the gradients’ quality. Specifically, we propose a privacy-preserving homomorphic noise mechanism in which the noise of each gradient sums to zero after aggregation, ensuring the availability of aggregated gradient. In addition, we design a joint audit algorithm that can locate malicious data owners without decrypting individual gradients. Through security analysis and experimental evaluation, we demonstrate that our approach can defend against malicious gradient attacks in federated learning.     

Stability assessment of 3D reinforced soil structures under steady unsaturated infiltration

Internal stability assessment of geosynthetic-reinforced soil structures (GRSSs) has been commonly carried out assuming plane-strain conditions and dry backfills. However, failures of GRSSs usually show three-dimensional (3D) features and occur under unsaturated conditions. A procedure based on the kinematic limit-analysis method is proposed herein to assess 3D effects and the role of steady unsaturated infiltration on the required geosynthetic strength for GRSSs. A suction stress-based framework is used to describe the soil stress behavior under steady unsaturated infiltration. Based on the principle of energy-work balance, the required geosynthetic strength is determined. A comparison analysis with the prior research is conducted to verify the developed method. Two kinds of backfills, i.e., high-quality backfill and marginal backfill, are considered for comparison in this work. It is shown that accounting for 3D effects and the role of unsaturated infiltration considerably reduces the required geosynthetic strength. The 3D effects are primarily affected by the width-to-height ratio of GRSSs, and the contribution of unsaturated infiltration is mainly influenced by the soil type, flow rate, GRSS's height, and location of the water table.     

A Tradeoff Between Accuracy and Speed for K-Means Seed Determination

With a sharp increase in the information volume, analyzing and retrieving this vast data volume is much more essential than ever. One of the main techniques that would be beneficial in this regard is called the Clustering method. Clustering aims to classify objects so that all objects within a cluster have similar features while other objects in different clusters are as distinct as possible. One of the most widely used clustering algorithms with the well and approved performance in different applications is the k-means algorithm. The main problem of the k-means algorithm is its performance which can be directly affected by the selection in the primary clusters. Lack of attention to this crucial issue has consequences such as creating empty clusters and decreasing the convergence time. Besides, the selection of appropriate initial seeds can reduce the cluster’s inconsistency. In this paper, we present a new method to determine the initial seeds of the k-mean algorithm to improve the accuracy and decrease the number of iterations of the algorithm. For this purpose, a new method is proposed considering the average distance between objects to determine the initial seeds. Our method attempts to provide a proper tradeoff between the accuracy and speed of the clustering algorithm. The experimental results showed that our proposed approach outperforms the Chithra with 1.7% and 2.1% in terms of clustering accuracy for Wine and Abalone detection data, respectively. Furthermore, achieved results indicate that comparing with the Reverse Nearest Neighbor (RNN) search approach, the proposed method has a higher convergence speed.     

Sustainable phenolic thermosets coatings derived from urushiol

The over-exploitation of finite fossil resources and/or the increased environmental and sustainable awareness inspire scientists and technologists to search for inexpensive alternatives from renewable chemicals. Phenol formaldehyde (PF) resins, the oldest type of synthetic polymers with good mechanical properties and heat resistance, are widely used in the production of coatings, laminates, molding compositions, and glues. Here, biobased urushiol-derived PF resins were synthesized from the alkali-catalyzed reaction between urushiol and formaldehyde. The chemical compositions and molecular structures of resole resins were characterized by carbon-13 nuclear magnetic resonance and Fourier transform infrared spectroscopy, and their curing behaviors were studied by differential scanning calorimetry. The as-prepared urushiol-derived resole resins had methylol (Ph−CH₂OH), ortho- and para-hemiformal groups (Ph−CH₂OCH₂OH), and the para−para/ortho−para/ortho−ortho links of methylene groups (Ph−CH₂−Ph), whereas the resole resins had low curing temperatures at about 100–113°C. Additionally, given the long side alkyl group moiety on the aromatic rings of urushiol, the films of cured urushiol-derived resole resins had low glass transition temperatures of 132 ± 2°C. Furthermore, the as-prepared urushiol-derived coatings exhibited excellent physical and mechanical properties.     

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。     

Main trends on the simulation of the drying of refractory castables - Review

Refractory castables develop microstructures after curing that behave as partially saturated porous media. Upon heating (during its drying stage), the steam generated by the physical and chemically bond water can result in pore pressurization and explosive spalling. Numerical modeling can provide guidelines for designing safer heat-up profiles and also a better understanding of the mechanisms that lead to catastrophic damage. This work aims to review the fundamentals and models available, providing insightful thoughts on the current trends of the drying phenomena of ceramic compositions. The review also highlights that there are models better oriented to result in reasonable predictions of pore pressure values and others focused on a more accurate representation of the main physical phenomena that take place during heating. According to the findings, there are still various challenges to attain accurate models with high applicability capable of yielding safer and more efficient drying of refractory castables.     

Toward High-Performance Dihydrophenazine-Based Conjugated Microporous Polymer Cathodes for Dual-Ion Batteries through Donor–Acceptor Structural Design

Recent studies have demonstrated that dihydrophenazine (Pz) with high redox-reversibility and high theoretical capacity is an attractive building block to construct p-type polymer cathodes for dual-ion batteries. However, most reported Pz-based polymer cathodes to date still suffer from low redox activity, slow kinetics, and short cycling life. Herein, a donor–acceptor (D–A) Pz-based conjugated microporous polymer (TzPz) cathode is constructed by integrating the electron-donating Pz unit and the electron-withdrawing 2,4,6-triphenyl-1,3,5-triazine (Tz) unit into a polymer chain. The D–A type structure enhances the polymer conjugation degree and decreases the band gap of TzPz, facilitating electron transportation along the polymer skeletons. Therefore the TzPz cathode for dual-ion battery shows a high reversible capacity of 192 mAh g⁻¹ at 0.2 A g⁻¹ with excellent rate performance (108 mAh g⁻¹ at 30 A g⁻¹), which is much higher than that of its counterpart polymer BzPz produced from 1,3,5-triphenylbenzene (Bz) and Pz (148 and 44 mAh g⁻¹ at 0.2 and 10 A g⁻¹, respectively). More importantly, the TzPz cathode also shows a long and stable cyclability of more than 10 000 cycles. These results demonstrate that the D–A structural design is an efficient strategy for developing high-performance polymer cathodes for dual-ion batteries.     

Residual attention-based tracking-by-detection network with attention-driven data augmentation

Tracking-by-detection (TBD) is a significant framework for visual object tracking. However, current trackers are usually updated online based on random sampling with a probability distribution. The performance of the learning-based TBD trackers is limited by the lack of discriminative features, especially when the background is full of semantic distractors. We propose an attention-driven data augmentation method, in which a residual attention mechanism is integrated into the TBD tracking network as supplementary references to identify discriminative image features. A mask generating network is used to simulate changes in target appearances to obtain positive samples, where attention information and image features are combined to identify discriminative features. In addition, we propose a method for mining hard negative samples, which searches for semantic distractors with the response of the attention module. The experiments on the OTB2015, UAV123, and LaSOT benchmarks show that this method achieves competitive performance in terms of accuracy and robustness.