Black-box hard-label adversarial attack on text is a practical and challenging task, as the text data space is inherently discrete and non-differentiable, and only the predicted label is accessible. Research on this problem is still in the embryonic stage and only a few methods are available. Nevertheless, existing methods rely on the complex heuristic algorithm or unreliable gradient estimation strategy, which probably fall into the local optimum and inevitably consume numerous queries, thus are difficult to craft satisfactory adversarial examples with high semantic similarity and low perturbation rate in a limited query budget. To alleviate above issues, we propose a simple yet effective framework to generate high quality textual adversarial examples under the black-box hard-label attack scenarios, named HQA-Attack.

Semi-Supervised Video Salient Object Detection (SS-VSOD) is challenging because of the lack of temporal information caused by sparse annotations in video sequences. Most works address this problem by generating pseudo labels for unlabeled data. However, error-prone pseudo labels negatively affect the VOSD model. Therefore, a deeper insight into pseudo labels should be developed. In this work, we aim to explore 1) how to utilize the incorrect predictions in pseudo labels to guide the network to generate more robust pseudo labels and 2) how to further screen out the noise that still exists in the improved pseudo labels. To this end, we propose an Uncertainty-Guided Pseudo Label Generator (UGPLG), which makes full use of inter-frame information to ensure the temporal consistency of the pseudo-labels and improves the robustness of the pseudo labels by strengthening the learning of difficult scenarios. Furthermore, we also introduce adversarial learning to address the noise problems in pseudo labels, guaranteeing the positive guidance of pseudo labels during model training. Experimental results demonstrate that our methods outperform existing semi-supervised method and partial fully-supervised methods across five public benchmarks of DAVIS, FBMS, MCL, ViSal, and SegTrack-V2. Code and dataset are available at https://github.com/Lanezzz/UGPL.

The aim of this study is to analyze the effect of serum metabolites on diabetic nephropathy (DN) and predict the prevalence of DN through a machine learning approach. The dataset consists of 548 patients from April 2018 to April 2019 in Second Affiliated Hospital of Dalian Medical University (SAHDMU). We select the optimal 38 features through a Least absolute shrinkage and selection operator (LASSO) regression model and a 10-fold cross-validation. We compare four machine learning algorithms, including eXtreme Gradient Boosting (XGB), random forest, decision tree and logistic regression, by AUC-ROC curves, decision curves, calibration curves. We quantify feature importance and interaction effects in the optimal predictive model by Shapley Additive exPlanations (SHAP) method. The XGB model has the best performance to screen for DN with the highest AUC value of 0.966. The XGB model also gains more clinical net benefits than others and the fitting degree is better. In addition, there are significant interactions between serum metabolites and duration of diabetes. We develop a predictive model by XGB algorithm to screen for DN. C2, C5DC, Tyr, Ser, Met, C24, C4DC, and Cys have great contribution in the model, and can possibly be biomarkers for DN.

Conventional robots based on separate joints do not always perform well in complex real-world tasks, particularly those that involve the dexterous manipulation of objects. Some roboticists have thus been trying to devise continuum robots, robotic platforms characterized by infinite degrees of freedom and no fixed number of joints. Continuum robots are typically based on cables or other deformable components that can move more freely and are not restricted by fixed joint structures. Despite these advantages, many continuum robot designs proposed still cannot yet efficiently navigate complex and unstructured environments. Researchers at Sun Yat-Sen University, Dalian University of Technology and London South Bank University have recently developed a new continuum robot inspired by the trunks of elephants.

DALIAN, China -- The race for supremacy in AI-powered automotive voice recognition is heating up in China, as the world's biggest auto market increasingly becomes a standard-bearer for technology.

The acquisition includes the Nand SSD business, the Nand component and wafer business, and the Dalian Nand memory manufacturing facility in China. Intel will however keep its distinct Intel Optane business. The companies hope to close the deal after receiving all of the necessary governmental approvals in late 2021. Under the agreement, SK hynix will first acquire the Nand SSD business (including NAND SSD-associated IP and employees), and the Dalian facility, with a first payment of USD 7 billion. SK hynix will acquire the remaining assets, including IP related to the manufacture and design of Nand flash wafers, R&D employees, and the Dalian fab workforce, upon final closing sometime in March 2025 with the remaining payment of USD 2 billion.

Vanadium flow batteries (VFBs) are promising for stationary large-scale energy storage due to their high safety, long cycle life, and high efficiency. The cost of a VFB system mainly depends on the VFB stack, electrolyte, and control system. Developing a VFB stack from lab to industrial scale can take years of experiments due to complex factors, from key materials to battery architecture. Novel methods to accurately predict the performance and cost of a VFB stack and further system are needed in order to accelerate the commercialization of VFBs. Recently, a research team led by Prof. Li Xianfeng from the Dalian Institute of Chemical Physics (DICP) of the Chinese Academy of Sciences proposed a machine learning-based strategy to predict and optimize the performance and cost of VFBs.

CNN are great blablabla… Let's get to the point. SOTA for image classification on Imagenet is EfficientNet with 88.5% top 1 accuracy in 2020. In this article, I introduce a combination of EfficientNet and Efficient Channel Attention (ECA) to highlight the results of the ECA paper from Tianjin/Dalian/Harbin universities. MobileNetV2 is composed of multiple blocks which are called linear bottlenecks or inverted residuals (they're almost the same). Linear Bottleneck is a residual layer composed of one 1x1 convolution, followed by a 3x3 depthwise convolution, then finally a 1x1 convolution.

But computer scientists have been working on alternative models, even as the public has grown weary of having their data used and abused. Dawn Song, a professor at the University of California, Berkeley, and one of the world's foremost experts in computer security and trustworthy artificial intelligence, envisions a new paradigm in which people control their data and are compensated for its use by corporations. While there have been many proposals for such a system, Professor Song is one actually building the platform to make it a reality. "As we talk about data as the new oil, it's particularly important to develop technologies that can utilize data in a privacy-preserving way," Professor Song said recently from her San Francisco office with an expansive view of the bay. It is an unlikely trajectory for Professor Song, who grew up in Dalian, China, a seaport in the northeastern province of Liaoning.

Robert Bosch Venture Capital has announced investment in AutoAI. The company is the developer and operator of intelligent connected vehicle service (CVS) for the new generation of autonomous vehicles. It is established as a separately run subsidiary of Navinfo and is involved in technological development, product development and service operations of intelligent navigation, CVS content, intelligent OS and solutions, CVS big data and operations. Based on the core mission of "Making Auto Intelligence Easier", the company has set up bases for front-end R&D and operations with a total of nearly 1,000 employees in Beijing, Shanghai, Shenzhen, Dalian and Shenyang. AUTOAI works jointly with automobile manufacturers, industry customers and ecosystem partners in the development of the next generation of autonomous driving, bring further cutting-edge technologies, leading products, ultimate experience and compassionate service to more partners, and aim to become an innovator and leader in the intelligent CVS industry.