如果不部署面部反动体相反的对策,则可以通过呈现印刷照片,视频或真实用户的硅面膜来欺骗面部识别系统。因此,面对表现攻击检测(PAD)在提供对数字设备的安全访问方面起着至关重要的作用。大多数现有的基于视频的垫子对策都无法应对视频中的远程时间变化。此外,在特征提取步骤之前的键框采样尚未在面部抗散热域中广泛研究。为了减轻这些问题,本文通过提出一种视频处理方案来提供一种数据采样方法,该方案基于高斯加权功能进行建模远程时间变化。具体而言,提出的方案将视频序列的连续T帧编码基于T帧的高斯加权总和,将视频序列的连续T帧编码为单个RGB图像。仅使用数据采样方案,我们证明可以在三个公共基准数据集的数据库内和数据库间测试方案中没有任何铃铛和哨子来实现最先进的性能;即,重播攻击,MSU-MFSD和CASIA-FASD。特别是,与跨数据库情景中的基线相比,该计划的误差(CASIA-FASD的15.2%至6.7%,重播攻击的5.9%至4.9%)。
translated by 谷歌翻译
面部表现攻击检测(PAD)在防御面部识别系统免受演示攻击方面起着重要作用。 PAD的成功很大程度上依赖于需要大量标记数据的监督学习,这对于视频尤其具有挑战性,通常需要专家知识。为了避免昂贵的标记数据收集,本文提出了一种通过运动预测进行自我监督视频表示学习的新方法。为了实现这一目标,我们基于三个RGB框架利用时间一致性,这些RGB帧在视频序列中以三个不同的时间获取。然后将获得的帧转换为灰度图像,其中每个图像被指定为三个不同的通道,例如R(红色),G(绿色)和B(蓝色),形成动态灰度片段(DGS)。由此激励,这些标签会自动生成,以通过使用视频的不同时间长度来基于DG的时间多样性,这对下游任务非常有帮助。从我们方法的自我监督性质中受益,我们报告了结果,结果表明,在四个公共基准数据集上的现有方法,即重播攻击,MSU-MFSD,CASIA-FASD和OULU-NPU。解释性分析是通过石灰和Grad-CAM技术进行的,以可视化DGS中使用的最重要功能。
translated by 谷歌翻译
Pre-trained transformers are now the de facto models in Natural Language Processing given their state-of-the-art results in many tasks and languages. However, most of the current models have been trained on languages for which large text resources are already available (such as English, French, Arabic, etc.). Therefore, there are still a number of low-resource languages that need more attention from the community. In this paper, we study the Algerian dialect which has several specificities that make the use of Arabic or multilingual models inappropriate. To address this issue, we collected more than one million Algerian tweets, and pre-trained the first Algerian language model: DziriBERT. When compared with existing models, DziriBERT achieves better results, especially when dealing with the Roman script. The obtained results show that pre-training a dedicated model on a small dataset (150 MB) can outperform existing models that have been trained on much more data (hundreds of GB). Finally, our model is publicly available to the community.
translated by 谷歌翻译
Continuous behavioural authentication methods add a unique layer of security by allowing individuals to verify their unique identity when accessing a device. Maintaining session authenticity is now feasible by monitoring users' behaviour while interacting with a mobile or Internet of Things (IoT) device, making credential theft and session hijacking ineffective. Such a technique is made possible by integrating the power of artificial intelligence and Machine Learning (ML). Most of the literature focuses on training machine learning for the user by transmitting their data to an external server, subject to private user data exposure to threats. In this paper, we propose a novel Federated Learning (FL) approach that protects the anonymity of user data and maintains the security of his data. We present a warmup approach that provides a significant accuracy increase. In addition, we leverage the transfer learning technique based on feature extraction to boost the models' performance. Our extensive experiments based on four datasets: MNIST, FEMNIST, CIFAR-10 and UMDAA-02-FD, show a significant increase in user authentication accuracy while maintaining user privacy and data security.
translated by 谷歌翻译
In this paper, we increase the availability and integration of devices in the learning process to enhance the convergence of federated learning (FL) models. To address the issue of having all the data in one location, federated learning, which maintains the ability to learn over decentralized data sets, combines privacy and technology. Until the model converges, the server combines the updated weights obtained from each dataset over a number of rounds. The majority of the literature suggested client selection techniques to accelerate convergence and boost accuracy. However, none of the existing proposals have focused on the flexibility to deploy and select clients as needed, wherever and whenever that may be. Due to the extremely dynamic surroundings, some devices are actually not available to serve as clients in FL, which affects the availability of data for learning and the applicability of the existing solution for client selection. In this paper, we address the aforementioned limitations by introducing an On-Demand-FL, a client deployment approach for FL, offering more volume and heterogeneity of data in the learning process. We make use of the containerization technology such as Docker to build efficient environments using IoT and mobile devices serving as volunteers. Furthermore, Kubernetes is used for orchestration. The Genetic algorithm (GA) is used to solve the multi-objective optimization problem due to its evolutionary strategy. The performed experiments using the Mobile Data Challenge (MDC) dataset and the Localfed framework illustrate the relevance of the proposed approach and the efficiency of the on-the-fly deployment of clients whenever and wherever needed with less discarded rounds and more available data.
translated by 谷歌翻译
将异常检测外包给第三方可以允许数据所有者克服资源限制(例如,在轻量级的IoT设备中),促进协作分析(例如,分布式或多方场景下的分布式或多方场景),并受益于较低的成本和专业知识(例如托管安全服务提供商)。尽管有这样的好处,但数据所有者可能会不愿外包异常检测而没有足够的隐私保护。为此,大多数现有的隐私解决方案将面临新的挑战,即保留隐私通常需要消除或减少数据条目之间的差异,而异常检测严重取决于该差异。最近,在本地分析设置下,通过将差异隐私(DP)保证的重点从“全部”到“良性”条目移动,这一冲突是在本地分析设置下解决的。在本文中,我们观察到这种方法不直接适用于外包设置,因为数据所有者在外包之前不知道哪些条目是“良性”的,因此无法选择地将DP应用于数据条目。因此,我们提出了一种新型的迭代解决方案,使数据所有者逐渐“脱离”良性条目的异常条目,以便第三方分析师可以通过足够的DP保证产生准确的异常结果。我们设计并实施了我们对异常检测(DPOAD)框架的差异私人外包,并通过从不同应用域中的真实数据进行实验,证明了其比基线拉普拉斯和无止痛机制的好处。
translated by 谷歌翻译
一阶逻辑的传统自动定理普通普及依赖于速度优化的搜索和许多手工制的启发式,旨在在各种域中工作。文献中的机器学习方法取决于这些传统的传统普通普通的自我或达到了可比性的比较短暂。在本文中,我们提出了一般的增量学习算法,用于培训域特定竞争员的一阶逻辑,而不是基本的给定 - 子算法,但使用学习的子句评分函数。子句被用作图表,并呈现给具有光谱特征的变压器网络。为了解决稀疏性和初步缺乏培训数据以及缺乏自然课程,我们适应后敏感经验重播到定理证明,即使没有找到证据。我们展示了这种方式培训的普通培训可以在TPTP数据集中匹配和有时会在证明的数量和质量方面超越TPTP数据集。
translated by 谷歌翻译
在本文中,我们解决了诱导的半监督学习问题,旨在获取样本数据的标签预测。所提出的方法称为最优传输诱导(OTI),有效地将最佳的传输基于传输的转换算法(OTP)扩展到二进制和多级设置的归纳任务。在多个数据集上进行一系列实验,以便将所提出的方法与最先进的方法进行比较。实验证明了我们方法的有效性。我们将我们的代码公开使用(代码可供选择:https://github.com/mouradelhamri/oti)。
translated by 谷歌翻译
在本文中,我们提出了一种对无监督域适应的新方法,与最佳运输,学习概率措施和无监督学习的概念相关。所提出的方法Hot-DA基于最佳运输的分层制定,其利用了由地面度量捕获的几何信息,源和目标域中的结构信息更丰富的结构信息。通过根据其类标签将样本分组到结构中,本质地形成标记的源域中的附加信息。在探索未标记的目标域中的隐藏结构的同时,通过Wassersein BaryCenter的学习概率措施的问题,我们证明是等同于光谱聚类。具有可控复杂性的玩具数据集的实验和两个具有挑战性的视觉适应数据集显示了所提出的方法的优越性。
translated by 谷歌翻译