ID 原文 译文
56158 最后列举了一些可微渲染可能的发展方向. At last, we point out some future research directions on differentiablerendering techniques.
56159 中小企业贷款在促进技术创新、推动经济发展、改善民生和增加就业等方面有着重要的作用. Small and medium-sized enterprises (SMEs) loans play an essential role in many aspects:
56160 为了满足商业银行的贷款评估标准,很多中小企业选择互相提供担保以获得授信,形成了结构复杂的担保网络. includingtechnological innovation, economic development, employment, and people’s livelihood, etc. In order to meetthe loan evaluation criteria of commercial banks, many SMEs choose to guarantee each other to obtain loans,thus forming a complex guarantee network.
56161 当借款方的贷款违约时,风险则沿着担保方向在网络中层层传播,由此造成的潜在系统性风险给国家的金融安全和监管带来了严峻的挑战. If the borrower defaults on the loan, the risk will be diffused to itsguarantors along with the contagion path, which may lead to systemic risk across the loan networks.
56162 因此,迫切需要发展相应的方法从系统角度对复杂金融担保网络中的传染路径进行风险评估和预测. This hasbrought severe challenges to the nation’s financial security and regulation. Thus, accurately rating the contagionpath is an urgent task for systematic risk management in the loan network. Therefore, we present a deep learning?based approach to the risk rating of contagion paths in the bank industry.
56163 本文提出了一种基于深度学习的风险评估模型,该方法应用图神经网络和注意力机制直接从网络化的贷款行为数据中学习风险特征,无需依赖于金融领域专业知识的人工特征工程. We leverage the graph neural networkand attention mechanism on graph-structured loan behavior to learn high-order representations, which do notrequire handcraft feature engineering.
56164 实验结果表明,本文设计的方法在多数评价指标上均优于现有的7个对比的基准模型.在传染路径风险评估任务中,比基准方法在精确率和召回率的调和平均数（F1-score）方面平均提升了2%15%.在新路径风险评估任务中,比最好的基准方法平均提升了3.5%. We demonstrate that our approach outperforms the existing baselineswith 2%∼15% improvements in risk rating and 3. 5% in the newly constructed path rating problem.
56165 结果表明了本文设计方法在传染路径风险评估中的有效性,可为监管部门和金融机构对担保网络进行系统性风险评估提供方法理论基础. The resultdemonstrates the effectiveness of our proposed approach, which provides an effective method and theory basis forregulatory commissions and financial institutions to monitor systematic risks in networked-guarantee loans.
56166 针对目标任务收集新类别的海量标注数据通常需要大量时间和人力成本,并已成为语义分割技术投入实际产业应用过程的主要瓶颈. Collecting a large amount of manually labeled training data is labor-intensive, thus often becomes themajor bottleneck when applying semantic segmentation techniques to real-world applications, especially for newcategories where no labeled data is available.
56167 本文旨在以"网络监督"的方式,在仅利用用户提供的目标类别关键词以及相应自动搜索到的网络数据的条件下实现语义分割模型的自主学习. In this paper, we aim at solving the problem of “webly-supervised”semantic segmentation relying purely on web searched images, where users only need to provide a single keywordfor each target category.