Convolutional neural networks (CNN) define the state-of-the-art solution on many perceptual tasks. However, current CNN approaches largely remain vulnerable against adversarial perturbations of the input that have been crafted specifically to fool the system while being quasi-imperceptible to the human eye. In recent years, various approaches have been proposed to defend CNNs against such attacks, for example by model hardening or by adding explicit defence mechanisms. Thereby, a small "detector" is included in the network and trained on the binary classification task of distinguishing genuine data from data containing adversarial perturbations. In this work, we propose a simple and light-weight detector, which leverages recent findings on the relation between networks' local intrinsic dimensionality (LID) and adversarial attacks. Based on a re-interpretation of the LID measure and several simple adaptations, we surpass the state-of-the-art on adversarial detection by a significant margin and reach almost perfect results in terms of F1-score for several networks and datasets. Sources available at: https://github.com/adverML/multiLID
translated by 谷歌翻译
Despite the success of convolutional neural networks (CNNs) in many academic benchmarks for computer vision tasks, their application in the real-world is still facing fundamental challenges. One of these open problems is the inherent lack of robustness, unveiled by the striking effectiveness of adversarial attacks. Current attack methods are able to manipulate the network's prediction by adding specific but small amounts of noise to the input. In turn, adversarial training (AT) aims to achieve robustness against such attacks and ideally a better model generalization ability by including adversarial samples in the trainingset. However, an in-depth analysis of the resulting robust models beyond adversarial robustness is still pending. In this paper, we empirically analyze a variety of adversarially trained models that achieve high robust accuracies when facing state-of-the-art attacks and we show that AT has an interesting side-effect: it leads to models that are significantly less overconfident with their decisions, even on clean data than non-robust models. Further, our analysis of robust models shows that not only AT but also the model's building blocks (like activation functions and pooling) have a strong influence on the models' prediction confidences. Data & Project website: https://github.com/GeJulia/robustness_confidences_evaluation
translated by 谷歌翻译
最近,变形金刚在图像分类中表现出巨大的潜力,并在ImageNet基准测试中建立了最先进的结果。然而,与CNN相比,变压器会缓慢收敛,并且由于缺乏空间电感偏见而容易过度拟合低数据。这种空间电感偏见可能特别有益,因为输入图像的2D结构在变压器中不能很好地保存。在这项工作中,我们提出了空间先验增强的自我注意力(SP-SA),这是为视觉变压器量身定制的香草自我注意力(SA)的新型变体。空间先验(SP)是我们提出的归纳偏见家族,突出了某些空间关系。与卷积归纳偏见不同,被迫专注于硬编码的地方区域,我们提出的SP是由模型本身学到的,并考虑了各种空间关系。具体而言,注意力评分是在每个头部都强调某些空间关系的重点,并且这种学识渊博的空间灶可以彼此互补。基于SP-SA,我们提出了SP-VIT家族,该家族始终优于其他具有相似GFLOPS或参数的VIT模型。我们最大的型号SP-VIT-L与以前的最新模型相比,参数数量降低了近50%(SP-VIT-L 150m VS 271M的CAIT-M-36)在所有Imagenet-1K模型中,在224x224训练,并在384x384分辨率上进行了微调,该分辨率带有额外的数据。
translated by 谷歌翻译
在过去的几年中,卷积神经网络(CNN)一直是广泛的计算机视觉任务中的主导神经架构。从图像和信号处理的角度来看,这一成功可能会令人惊讶,因为大多数CNN的固有空间金字塔设计显然违反了基本的信号处理法,即在其下采样操作中对定理进行采样。但是,由于不良的采样似乎不影响模型的准确性,因此在模型鲁棒性开始受到更多关注之前,该问题已被广泛忽略。最近的工作[17]在对抗性攻击和分布变化的背景下,毕竟表明,CNN的脆弱性与不良下降采样操作引起的混叠伪像之间存在很强的相关性。本文以这些发现为基础,并引入了一个可混合的免费下采样操作,可以轻松地插入任何CNN体系结构:频lowcut池。我们的实验表明,结合简单而快速的FGSM对抗训练,我们的超参数无操作员显着提高了模型的鲁棒性,并避免了灾难性的过度拟合。
translated by 谷歌翻译
在最近,对表现良好的神经体系结构(NAS)的高效,自动化的搜索引起了人们的关注。因此,主要的研究目标是减少对神经体系结构进行昂贵评估的必要性,同时有效地探索大型搜索空间。为此,替代模型将体系结构嵌入了潜在的空间并预测其性能,而神经体系结构的生成模型则可以在生成器借鉴的潜在空间内基于优化的搜索。替代模型和生成模型都具有促进结构良好的潜在空间中的查询搜索。在本文中,我们通过利用有效的替代模型和生成设计的优势来进一步提高查询效率和有前途的建筑生成之间的权衡。为此,我们提出了一个与替代预测指标配对的生成模型,该模型迭代地学会了从越来越有希望的潜在子空间中生成样品。这种方法可导致非常有效和高效的架构搜索,同时保持查询量较低。此外,我们的方法允许以一种直接的方式共同优化准确性和硬件延迟等多个目标。我们展示了这种方法的好处,不仅是W.R.T.优化体系结构以提高最高分类精度,但在硬件约束和在单个NAS基准测试中的最新方法和多个目标的最先进方法的优化。我们还可以在Imagenet上实现最先进的性能。该代码可在http://github.com/jovitalukasik/ag-net上找到。
translated by 谷歌翻译
最小成本多型问题(MP)是一种流行的方式,用于通过优化边缘成本优化二进制边缘标签来获取图形分解。虽然来自每个边缘的独立估计成本的MP的配方非常灵活,并且求解MP是NP - 硬度和较昂贵的。作为一个补救措施,最近的工作提出通过在预测过程中结合周期约束来预测对潜在冲突的认识来预测边缘概率。我们认为这种制定,同时为最终到最终学习的边缘重量提供第一步,是次优的,因为它建立在MP的松散松弛时。因此,我们提出了一种自适应CRF,允许逐步考虑更违反的限制,并因此地发出具有更高有效性的解决方案。对自然图像分割的BSDS500基准以及电子显微录制的实验表明,我们的方法产生了更精确的边缘检测和图像分割。
translated by 谷歌翻译
最近,Robustbench(Croce等人2020)已成为图像分类网络的对抗鲁棒性的广泛认可的基准。在其最常见的子任务中,Robustbench评估并在Autactack(CRoce和Hein 2020b)下的Cifar10上的培训神经网络的对抗性鲁棒性与L-Inf Perturnations限制在EPS = 8/255中。对于目前最佳表演模型的主要成绩约为60%的基线,这是为了表征这项基准是非常具有挑战性的。尽管最近的文献普遍接受,我们的目标是促进讨论抢劫案作为鲁棒性的关键指标的讨论,这可能是广泛化的实际应用。我们的论证与这篇文章有两倍,并通过本文提出过多的实验支持:我们认为i)通过ICATACK与L-INF的数据交替,EPS = 8/255是不切实际的强烈的,导致完美近似甚至通过简单的检测算法和人类观察者的对抗性样本的检测速率。我们还表明,其他攻击方法更难检测,同时实现类似的成功率。 ii)在CIFAR10这样的低分辨率数据集上导致低分辨率数据集不概括到更高的分辨率图像,因为基于梯度的攻击似乎与越来越多的分辨率变得更加可检测。
translated by 谷歌翻译
最近,对AutoAtack(Croce和Hein,2020B)框架对图像分类网络的对抗攻击已经引起了很多关注。虽然AutoAtactack显示了非常高的攻击成功率,但大多数防御方法都专注于网络硬化和鲁棒性增强,如对抗性培训。这样,目前最佳报告的方法可以承受约66%的CIFAR10对抗的例子。在本文中,我们研究了自动攻击的空间和频域属性,并提出了替代防御。在推理期间,我们检测到对抗性攻击而不是硬化网络,而不是硬化网络,而不是硬化网络。基于频域中的相当简单和快速的分析,我们介绍了两种不同的检测算法。首先,黑匣子检测器只在输入图像上运行,在两种情况下,在AutoAtack Cifar10基准测试中获得100%的检测精度,并且在ImageNet上为99.3%。其次,使用CNN特征图的分析的白箱检测器,在相同的基准上的检出率也为100%和98.7%。
translated by 谷歌翻译
The FlowNet demonstrated that optical flow estimation can be cast as a learning problem. However, the state of the art with regard to the quality of the flow has still been defined by traditional methods. Particularly on small displacements and real-world data, FlowNet cannot compete with variational methods. In this paper, we advance the concept of end-to-end learning of optical flow and make it work really well. The large improvements in quality and speed are caused by three major contributions: first, we focus on the training data and show that the schedule of presenting data during training is very important. Second, we develop a stacked architecture that includes warping of the second image with intermediate optical flow. Third, we elaborate on small displacements by introducing a subnetwork specializing on small motions. FlowNet 2.0 is only marginally slower than the original FlowNet but decreases the estimation error by more than 50%. It performs on par with state-of-the-art methods, while running at interactive frame rates. Moreover, we present faster variants that allow optical flow computation at up to 140fps with accuracy matching the original FlowNet.
translated by 谷歌翻译
气溶胶颗粒通过吸收和散射辐射并影响云特性在气候系统中起重要作用。它们也是气候建模的最大不确定性来源之一。由于计算限制,许多气候模型不包括足够详细的气溶胶。为了表示关键过程,必须考虑气雾微物理特性和过程。这是在使用M7 Microphysics的Echam-Ham全球气候气溶胶模型中完成的,但是高计算成本使得以更精细的分辨率或更长的时间运行非常昂贵。我们的目标是使用机器学习以足够的准确性模仿微物理学模型,并通过在推理时间快速降低计算成本。原始M7模型用于生成输入输出对的数据以训练其上的神经网络。我们能够学习变量的平均$ r^2 $得分为$ 77.1 \%$ $。我们进一步探讨了用物理知识为神经网络提供信息和限制的方法,以减少群众侵犯并实施质量积极性。与原始型号相比,在GPU上,我们达到了高达64倍的加速。
translated by 谷歌翻译