Neural networks are prone to catastrophic forgetting when trained incrementally on different tasks. Popular incremental learning methods mitigate such forgetting by retaining a subset of previously seen samples and replaying them during the training on subsequent tasks. However, this is not always possible, e.g., due to data protection regulations. In such restricted scenarios, one can employ generative models to replay either artificial images or hidden features to a classifier. In this work, we propose Genifer (GENeratIve FEature-driven image Replay), where a generative model is trained to replay images that must induce the same hidden features as real samples when they are passed through the classifier. Our technique therefore incorporates the benefits of both image and feature replay, i.e.: (1) unlike conventional image replay, our generative model explicitly learns the distribution of features that are relevant for classification; (2) in contrast to feature replay, our entire classifier remains trainable; and (3) we can leverage image-space augmentations, which increase distillation performance while also mitigating overfitting during the training of the generative model. We show that Genifer substantially outperforms the previous state of the art for various settings on the CIFAR-100 and CUB-200 datasets.
translated by 谷歌翻译
Deep learning models can achieve high accuracy when trained on large amounts of labeled data. However, real-world scenarios often involve several challenges: Training data may become available in installments, may originate from multiple different domains, and may not contain labels for training. Certain settings, for instance medical applications, often involve further restrictions that prohibit retention of previously seen data due to privacy regulations. In this work, to address such challenges, we study unsupervised segmentation in continual learning scenarios that involve domain shift. To that end, we introduce GarDA (Generative Appearance Replay for continual Domain Adaptation), a generative-replay based approach that can adapt a segmentation model sequentially to new domains with unlabeled data. In contrast to single-step unsupervised domain adaptation (UDA), continual adaptation to a sequence of domains enables leveraging and consolidation of information from multiple domains. Unlike previous approaches in incremental UDA, our method does not require access to previously seen data, making it applicable in many practical scenarios. We evaluate GarDA on two datasets with different organs and modalities, where it substantially outperforms existing techniques.
translated by 谷歌翻译
Despite significant advances, the performance of state-of-the-art continual learning approaches hinges on the unrealistic scenario of fully labeled data. In this paper, we tackle this challenge and propose an approach for continual semi-supervised learning -- a setting where not all the data samples are labeled. An underlying issue in this scenario is the model forgetting representations of unlabeled data and overfitting the labeled ones. We leverage the power of nearest-neighbor classifiers to non-linearly partition the feature space and learn a strong representation for the current task, as well as distill relevant information from previous tasks. We perform a thorough experimental evaluation and show that our method outperforms all the existing approaches by large margins, setting a strong state of the art on the continual semi-supervised learning paradigm. For example, on CIFAR100 we surpass several others even when using at least 30 times less supervision (0.8% vs. 25% of annotations).
translated by 谷歌翻译
基于正规化的方法有利于缓解类渐进式学习中的灾难性遗忘问题。由于缺乏旧任务图像,如果分类器在新图像上产生类似的输出,它们通常会假设旧知识得到很好的保存。在本文中,我们发现他们的效果很大程度上取决于旧课程的性质:它们在彼此之间容易区分的课程上工作,但可能在更细粒度的群体上失败,例如,男孩和女孩。在SPIRIT中,此类方法将新数据项目投入到完全连接层中的权重向量中跨越的特征空间,对应于旧类。由此产生的预测在细粒度的旧课程上是相似的,因此,新分类器将逐步失去这些课程的歧视能力。为了解决这个问题,我们提出了一种无记忆生成的重播策略,通过直接从旧分类器生成代表性的旧图像并结合新的分类器培训的新数据来保留细粒度的旧阶级特征。为了解决所产生的样本的均化问题,我们还提出了一种分集体损失,使得产生的样品之间的Kullback Leibler(KL)发散。我们的方法最好是通过先前的基于正规化的方法补充,证明是为了易于区分的旧课程有效。我们验证了上述关于CUB-200-2011,CALTECH-101,CIFAR-100和微小想象的设计和见解,并表明我们的策略优于现有的无记忆方法,并具有清晰的保证金。代码可在https://github.com/xmengxin/mfgr获得
translated by 谷歌翻译
很少有人提出了几乎没有阶级的课程学习(FSCIL),目的是使深度学习系统能够逐步学习有限的数据。最近,一位先驱声称,通常使用的基于重播的课堂学习方法(CIL)是无效的,因此对于FSCIL而言并不是首选。如果真理,这对FSCIL领域产生了重大影响。在本文中,我们通过经验结果表明,采用数据重播非常有利。但是,存储和重播旧数据可能会导致隐私问题。为了解决此问题,我们或建议使用无数据重播,该重播可以通过发电机综合数据而无需访问真实数据。在观察知识蒸馏的不确定数据的有效性时,我们在发电机培训中强加了熵正则化,以鼓励更不确定的例子。此外,我们建议使用单速样标签重新标记生成的数据。这种修改使网络可以通过完全减少交叉渗透损失来学习,从而减轻了在常规知识蒸馏方法中平衡不同目标的问题。最后,我们对CIFAR-100,Miniimagenet和Cub-200展示了广泛的实验结果和分析,以证明我们提出的效果。
translated by 谷歌翻译
大多数元学习方法都假设存在于可用于基本知识的情节元学习的一组非常大的标记数据。这与更现实的持续学习范例形成对比,其中数据以包含不相交类的任务的形式逐步到达。在本文中,我们考虑了这个增量元学习(IML)的这个问题,其中类在离散任务中逐步呈现。我们提出了一种方法,我们调用了IML,我们称之为eCISODIC重播蒸馏(ERD),该方法将来自当前任务的类混合到当前任务中,当研究剧集时,来自先前任务的类别示例。然后将这些剧集用于知识蒸馏以最大限度地减少灾难性的遗忘。四个数据集的实验表明ERD超越了最先进的。特别是,在一次挑战的单次次数较挑战,长任务序列增量元学习场景中,我们将IML和联合训练与当前状态的3.5%/ 10.1%/ 13.4%之间的差距降低我们在Diered-ImageNet / Mini-ImageNet / CIFAR100上分别为2.6%/ 2.9%/ 5.0%。
translated by 谷歌翻译
人类智慧的主食是以不断的方式获取知识的能力。在Stark对比度下,深网络忘记灾难性,而且为此原因,类增量连续学习促进方法的子字段逐步学习一系列任务,将顺序获得的知识混合成综合预测。这项工作旨在评估和克服我们以前提案黑暗体验重播(Der)的陷阱,这是一种简单有效的方法,将排练和知识蒸馏结合在一起。灵感来自于我们的思想不断重写过去的回忆和对未来的期望,我们赋予了我的能力,即我的能力来修改其重播记忆,以欢迎有关过去数据II的新信息II)为学习尚未公开的课程铺平了道路。我们表明,这些策略的应用导致了显着的改进;实际上,得到的方法 - 被称为扩展-DAR(X-DER) - 优于标准基准(如CiFar-100和MiniimAgeNet)的技术状态,并且这里引入了一个新颖的。为了更好地了解,我们进一步提供了广泛的消融研究,以证实并扩展了我们以前研究的结果(例如,在持续学习设置中知识蒸馏和漂流最小值的价值)。
translated by 谷歌翻译
Although deep learning approaches have stood out in recent years due to their state-of-the-art results, they continue to suffer from catastrophic forgetting, a dramatic decrease in overall performance when training with new classes added incrementally. This is due to current neural network architectures requiring the entire dataset, consisting of all the samples from the old as well as the new classes, to update the model-a requirement that becomes easily unsustainable as the number of classes grows. We address this issue with our approach to learn deep neural networks incrementally, using new data and only a small exemplar set corresponding to samples from the old classes. This is based on a loss composed of a distillation measure to retain the knowledge acquired from the old classes, and a cross-entropy loss to learn the new classes. Our incremental training is achieved while keeping the entire framework end-to-end, i.e., learning the data representation and the classifier jointly, unlike recent methods with no such guarantees. We evaluate our method extensively on the CIFAR-100 and Im-ageNet (ILSVRC 2012) image classification datasets, and show state-of-the-art performance.
translated by 谷歌翻译
Artificial neural networks thrive in solving the classification problem for a particular rigid task, acquiring knowledge through generalized learning behaviour from a distinct training phase. The resulting network resembles a static entity of knowledge, with endeavours to extend this knowledge without targeting the original task resulting in a catastrophic forgetting. Continual learning shifts this paradigm towards networks that can continually accumulate knowledge over different tasks without the need to retrain from scratch. We focus on task incremental classification, where tasks arrive sequentially and are delineated by clear boundaries. Our main contributions concern (1) a taxonomy and extensive overview of the state-of-the-art; (2) a novel framework to continually determine the stability-plasticity trade-off of the continual learner; (3) a comprehensive experimental comparison of 11 state-of-the-art continual learning methods and 4 baselines. We empirically scrutinize method strengths and weaknesses on three benchmarks, considering Tiny Imagenet and large-scale unbalanced iNaturalist and a sequence of recognition datasets. We study the influence of model capacity, weight decay and dropout regularization, and the order in which the tasks are presented, and qualitatively compare methods in terms of required memory, computation time and storage.
translated by 谷歌翻译
Continual Learning (CL) is a field dedicated to devise algorithms able to achieve lifelong learning. Overcoming the knowledge disruption of previously acquired concepts, a drawback affecting deep learning models and that goes by the name of catastrophic forgetting, is a hard challenge. Currently, deep learning methods can attain impressive results when the data modeled does not undergo a considerable distributional shift in subsequent learning sessions, but whenever we expose such systems to this incremental setting, performance drop very quickly. Overcoming this limitation is fundamental as it would allow us to build truly intelligent systems showing stability and plasticity. Secondly, it would allow us to overcome the onerous limitation of retraining these architectures from scratch with the new updated data. In this thesis, we tackle the problem from multiple directions. In a first study, we show that in rehearsal-based techniques (systems that use memory buffer), the quantity of data stored in the rehearsal buffer is a more important factor over the quality of the data. Secondly, we propose one of the early works of incremental learning on ViTs architectures, comparing functional, weight and attention regularization approaches and propose effective novel a novel asymmetric loss. At the end we conclude with a study on pretraining and how it affects the performance in Continual Learning, raising some questions about the effective progression of the field. We then conclude with some future directions and closing remarks.
translated by 谷歌翻译
恶意软件(恶意软件)分类为持续学习(CL)制度提供了独特的挑战,这是由于每天收到的新样本的数量以及恶意软件的发展以利用新漏洞。在典型的一天中,防病毒供应商将获得数十万个独特的软件,包括恶意和良性,并且在恶意软件分类器的一生中,有超过十亿个样品很容易积累。鉴于问题的规模,使用持续学习技术的顺序培训可以在减少培训和存储开销方面提供可观的好处。但是,迄今为止,还没有对CL应用于恶意软件分类任务的探索。在本文中,我们研究了11种应用于三个恶意软件任务的CL技术,涵盖了常见的增量学习方案,包括任务,类和域增量学习(IL)。具体而言,使用两个现实的大规模恶意软件数据集,我们评估了CL方法在二进制恶意软件分类(domain-il)和多类恶意软件家庭分类(Task-IL和类IL)任务上的性能。令我们惊讶的是,在几乎所有情况下,持续的学习方法显着不足以使训练数据的幼稚关节重播 - 在某些情况下,将精度降低了70个百分点以上。与关节重播相比,有选择性重播20%的存储数据的一种简单方法可以实现更好的性能,占训练时间的50%。最后,我们讨论了CL技术表现出乎意料差的潜在原因,希望它激发进一步研究在恶意软件分类域中更有效的技术。
translated by 谷歌翻译
这项工作调查了持续学习(CL)与转移学习(TL)之间的纠缠。特别是,我们阐明了网络预训练的广泛应用,强调它本身受到灾难性遗忘的影响。不幸的是,这个问题导致在以后任务期间知识转移的解释不足。在此基础上,我们提出了转移而不忘记(TWF),这是在固定的经过预定的兄弟姐妹网络上建立的混合方法,该方法不断传播源域中固有的知识,通过层次损失项。我们的实验表明,TWF在各种设置上稳步优于其他CL方法,在各种数据集和不同的缓冲尺寸上,平均每种类型的精度增长了4.81%。
translated by 谷歌翻译
在本文中,我们为连续表示学习问题提出了一种新颖的培训程序,其中依次学习了神经网络模型,以减轻视觉搜索任务中的灾难性遗忘。我们的方法称为对比度有监督的蒸馏(CSD),在学习判别特征的同时,还会减少忘记。这是通过在蒸馏设置中利用标签信息来实现的,在蒸馏设置中,从教师模型中对学生模型进行了相反的学习。广泛的实验表明,CSD在减轻灾难性遗忘方面的表现优于当前最新方法。我们的结果还提供了进一步的证据,表明在视觉检索任务中评估的功能忘记不像分类任务那样灾难性。代码:https://github.com/niccobiondi/contrastivesupervisedistillation。
translated by 谷歌翻译
当自我监督的模型已经显示出比在规模上未标记的数据训练的情况下的监督对方的可比视觉表现。然而,它们的功效在持续的学习(CL)场景中灾难性地减少,其中数据被顺序地向模型呈现给模型。在本文中,我们表明,通过添加将表示的当前状态映射到其过去状态,可以通过添加预测的网络来无缝地转换为CL的蒸馏机制。这使我们能够制定一个持续自我监督的视觉表示的框架,学习(i)显着提高了学习象征的质量,(ii)与若干最先进的自我监督目标兼容(III)几乎没有近似参数调整。我们通过在各种CL设置中培训六种受欢迎的自我监督模型来证明我们的方法的有效性。
translated by 谷歌翻译
最近的自我监督学习方法能够学习高质量的图像表示,并通过监督方法关闭差距。但是,这些方法无法逐步获取新的知识 - 事实上,它们实际上主要仅用为具有IID数据的预训练阶段。在这项工作中,我们在没有额外的记忆或重放的情况下调查持续学习制度的自我监督方法。为防止忘记以前的知识,我们提出了功能正规化的使用。我们将表明,朴素的功能正则化,也称为特征蒸馏,导致可塑性的低可塑性,因此严重限制了连续的学习性能。为了解决这个问题,我们提出了预测的功能正则化,其中一个单独的投影网络确保新学习的特征空间保留了先前的特征空间的信息,同时允许学习新功能。这使我们可以防止在保持学习者的可塑性时忘记。针对应用于自我监督的其他增量学习方法的评估表明我们的方法在不同场景和多个数据集中获得竞争性能。
translated by 谷歌翻译
受到正规彩票假说(RLTH)的启发,该假说假设在密集网络中存在平稳(非二进制)子网,以实现密集网络的竞争性能,我们提出了几个播放类增量学习(FSCIL)方法。 to as \ emph {soft-subnetworks(softnet)}。我们的目标是逐步学习一系列会议,每个会议在每个课程中只包含一些培训实例,同时保留了先前学到的知识。软网络在基本训练会议上共同学习模型权重和自适应非二进制软面具,每个面具由主要和次要子网组成;前者的目的是最大程度地减少训练期间的灾难性遗忘,而后者的目的是避免在每个新培训课程中过度拟合一些样本。我们提供了全面的经验验证,表明我们的软网络通过超越基准数据集的最先进基准的性能来有效地解决了几个弹药的学习问题。
translated by 谷歌翻译
Anomaly Detection is a relevant problem that arises in numerous real-world applications, especially when dealing with images. However, there has been little research for this task in the Continual Learning setting. In this work, we introduce a novel approach called SCALE (SCALing is Enough) to perform Compressed Replay in a framework for Anomaly Detection in Continual Learning setting. The proposed technique scales and compresses the original images using a Super Resolution model which, to the best of our knowledge, is studied for the first time in the Continual Learning setting. SCALE can achieve a high level of compression while maintaining a high level of image reconstruction quality. In conjunction with other Anomaly Detection approaches, it can achieve optimal results. To validate the proposed approach, we use a real-world dataset of images with pixel-based anomalies, with the scope to provide a reliable benchmark for Anomaly Detection in the context of Continual Learning, serving as a foundation for further advancements in the field.
translated by 谷歌翻译
本文研究持续学习(CL)的逐步学习(CIL)。已经提出了许多方法来处理CIL中的灾难性遗忘(CF)。大多数方法都会为单个头网络中所有任务的所有类别构建单个分类器。为了防止CF,一种流行的方法是记住以前任务中的少数样本,并在培训新任务时重播它们。但是,这种方法仍然患有严重的CF,因为在内存中仅使用有限的保存样本数量来更新或调整了先前任务的参数。本文提出了一种完全不同的方法,该方法使用变压器网络为每个任务(称为多头模型)构建一个单独的分类器(头部),称为更多。与其在内存中使用保存的样本在现有方法中更新以前的任务/类的网络,不如利用保存的样本来构建特定任务分类器(添加新的分类头),而无需更新用于先前任务/类的网络。新任务的模型经过培训,可以学习任务的类别,并且还可以检测到不是从相同数据分布(即,均分布(OOD))的样本。这使测试实例属于的任务的分类器能够为正确的类产生高分,而其他任务的分类器可以产生低分,因为测试实例不是来自这些分类器的数据分布。实验结果表明,更多的表现优于最先进的基线,并且自然能够在持续学习环境中进行OOD检测。
translated by 谷歌翻译
持续学习(CL)旨在制定模仿人类能力顺序学习新任务的能力,同时能够保留从过去经验获得的知识。在本文中,我们介绍了内存约束在线连续学习(MC-OCL)的新问题,这对存储器开销对可能算法可以用于避免灾难性遗忘的记忆开销。最多,如果不是全部,之前的CL方法违反了这些约束,我们向MC-OCL提出了一种算法解决方案:批量蒸馏(BLD),基于正则化的CL方法,有效地平衡了稳定性和可塑性,以便学习数据流,同时保留通过蒸馏解决旧任务的能力。我们在三个公开的基准测试中进行了广泛的实验评估,经验证明我们的方法成功地解决了MC-OCL问题,并实现了需要更高内存开销的先前蒸馏方法的可比准确性。
translated by 谷歌翻译
人类的持续学习(CL)能力与稳定性与可塑性困境密切相关,描述了人类如何实现持续的学习能力和保存的学习信息。自发育以来,CL的概念始终存在于人工智能(AI)中。本文提出了对CL的全面审查。与之前的评论不同,主要关注CL中的灾难性遗忘现象,本文根据稳定性与可塑性机制的宏观视角来调查CL。类似于生物对应物,“智能”AI代理商应该是I)记住以前学到的信息(信息回流); ii)不断推断新信息(信息浏览:); iii)转移有用的信息(信息转移),以实现高级CL。根据分类学,评估度量,算法,应用以及一些打开问题。我们的主要贡献涉及I)从人工综合情报层面重新检查CL; ii)在CL主题提供详细和广泛的概述; iii)提出一些关于CL潜在发展的新颖思路。
translated by 谷歌翻译