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.

Modern machine learning suffers from catastrophic forgetting when learning new classes incrementally. The performance dramatically degrades due to the missing data of old classes. Incremental learning methods have been proposed to retain the knowledge acquired from the old classes, by using knowledge distilling and keeping a few exemplars from the old classes. However, these methods struggle to scale up to a large number of classes. We believe this is because of the combination of two factors: (a) the data imbalance between the old and new classes, and (b) the increasing number of visually similar classes. Distinguishing between an increasing number of visually similar classes is particularly challenging, when the training data is unbalanced. We propose a simple and effective method to address this data imbalance issue. We found that the last fully connected layer has a strong bias towards the new classes, and this bias can be corrected by a linear model. With two bias parameters, our method performs remarkably well on two large datasets: ImageNet (1000 classes) and MS-Celeb-1M (10000 classes), outperforming the state-of-the-art algorithms by 11.1% and 13.2% respectively.

A major open problem on the road to artificial intelligence is the development of incrementally learning systems that learn about more and more concepts over time from a stream of data. In this work, we introduce a new training strategy, iCaRL, that allows learning in such a classincremental way: only the training data for a small number of classes has to be present at the same time and new classes can be added progressively.iCaRL learns strong classifiers and a data representation simultaneously. This distinguishes it from earlier works that were fundamentally limited to fixed data representations and therefore incompatible with deep learning architectures. We show by experiments on CIFAR-100 and ImageNet ILSVRC 2012 data that iCaRL can learn many classes incrementally over a long period of time where other strategies quickly fail.

Conventionally, deep neural networks are trained offline, relying on a large dataset prepared in advance. This paradigm is often challenged in real-world applications, e.g. online services that involve continuous streams of incoming data. Recently, incremental learning receives increasing attention, and is considered as a promising solution to the practical challenges mentioned above. However, it has been observed that incremental learning is subject to a fundamental difficulty -catastrophic forgetting, namely adapting a model to new data often results in severe performance degradation on previous tasks or classes. Our study reveals that the imbalance between previous and new data is a crucial cause to this problem. In this work, we develop a new framework for incrementally learning a unified classifier, i.e. a classifier that treats both old and new classes uniformly. Specifically, we incorporate three components, cosine normalization, less-forget constraint, and inter-class separation, to mitigate the adverse effects of the imbalance. Experiments show that the proposed method can effectively rebalance the training process, thus obtaining superior performance compared to the existing methods. On CIFAR-100 and ImageNet, our method can reduce the classification errors by more than 6% and 13% respectively, under the incremental setting of 10 phases.

当在新的类或新任务上逐步训练时，深度神经网络易于灾难性遗忘，因为对新数据的适应导致旧课程和任务的性能急剧下降。通过使用小记忆进行排练和知识蒸馏，已证明最近的方法可有效缓解灾难性的遗忘。然而，由于内存的尺寸有限，旧的和新类可用的数据量之间的大不平衡仍然存在，这导致模型的整体精度恶化。为了解决这个问题，我们建议使用平衡的软制跨熵损失，并表明它可以与进出的方法相结合，以便在某些情况下降低培训过程的计算成本，以提高其性能。对竞争的想象，Subimagenet和CiFar100数据集的实验显示了最艺术态度的结果。

虽然灾难性遗忘的概念是直截了当的，但缺乏对其原因的研究。在本文中，我们系统地探索并揭示了课堂增量学习中灾难性遗忘的三个原因（CIL）。从代表学习的角度来看，（i）当学习者未能正确对准相同相位数据时，逐步忘记在训练所得和（ii）当学习者混淆当前相数据时发生相互相互混淆上一阶段。从特定于任务特定的角度来看，CIL模型遭受了（iii）分类器偏差的问题。在调查现有策略后，我们观察到缺乏关于如何防止相互局部混淆的研究。要启动对该具体问题的研究，我们提出了一种简单但有效的框架，CIL（C4IL）的对比阶级浓度。我们的框架利用了对比度学习的阶级集中效应，产生了具有更好的级别的紧凑性和阶级间可分离的表示分布。经验上，我们观察到C4IL显着降低了相互相连的概率，并且结果提高了多个数据集的多个CIL设置的性能。

本文在课堂增量学习中使用视觉变压器（VIT）研究。令人惊讶的是，天真地应用Vit替代卷积神经网络（CNNS）导致性能下降。我们的分析揭示了三个天然使用VIT的问题：（a）vit在课程中较小时具有非常缓慢的会聚，（b）在比CNN的模型中观察到新类的更多偏差，并且（c）适当的学习率Vit太低，无法学习良好的分类器。基于此分析，我们展示了这些问题可以简单地通过使用现有技术来解决：使用卷积杆，平衡FineTuning来纠正偏置，以及分类器的更高学习率。我们的简单解决方案名为Vitil（Vit用于增量学习），为所有三类增量学习设置实现了全新的最先进的保证金，为研究界提供了强大的基线。例如，在ImageNet-1000上，我们的体内体达到69.20％的前1个精度为500个初始类别的15个初始类别，5个增量步骤（每次100个新类），表现优于leulir + dde ​​1.69％。对于10个增量步骤（100个新课程）的更具挑战性的协议，我们的方法优于PODNet 7.27％（65.13％与57.86％）。

Many modern computer vision algorithms suffer from two major bottlenecks: scarcity of data and learning new tasks incrementally. While training the model with new batches of data the model looses it's ability to classify the previous data judiciously which is termed as catastrophic forgetting. Conventional methods have tried to mitigate catastrophic forgetting of the previously learned data while the training at the current session has been compromised. The state-of-the-art generative replay based approaches use complicated structures such as generative adversarial network (GAN) to deal with catastrophic forgetting. Additionally, training a GAN with few samples may lead to instability. In this work, we present a novel method to deal with these two major hurdles. Our method identifies a better embedding space with an improved contrasting loss to make classification more robust. Moreover, our approach is able to retain previously acquired knowledge in the embedding space even when trained with new classes. We update previous session class prototypes while training in such a way that it is able to represent the true class mean. This is of prime importance as our classification rule is based on the nearest class mean classification strategy. We have demonstrated our results by showing that the embedding space remains intact after training the model with new classes. We showed that our method preformed better than the existing state-of-the-art algorithms in terms of accuracy across different sessions.

The dynamic expansion architecture is becoming popular in class incremental learning, mainly due to its advantages in alleviating catastrophic forgetting. However, task confusion is not well assessed within this framework, e.g., the discrepancy between classes of different tasks is not well learned (i.e., inter-task confusion, ITC), and certain priority is still given to the latest class batch (i.e., old-new confusion, ONC). We empirically validate the side effects of the two types of confusion. Meanwhile, a novel solution called Task Correlated Incremental Learning (TCIL) is proposed to encourage discriminative and fair feature utilization across tasks. TCIL performs a multi-level knowledge distillation to propagate knowledge learned from old tasks to the new one. It establishes information flow paths at both feature and logit levels, enabling the learning to be aware of old classes. Besides, attention mechanism and classifier re-scoring are applied to generate more fair classification scores. We conduct extensive experiments on CIFAR100 and ImageNet100 datasets. The results demonstrate that TCIL consistently achieves state-of-the-art accuracy. It mitigates both ITC and ONC, while showing advantages in battle with catastrophic forgetting even no rehearsal memory is reserved.

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).

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.

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。

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

我们研究了类新型小说类发现的新任务（class-incd），该任务是指在未标记的数据集中发现新型类别的问题，该问题通过利用已在包含脱节的标签数据集上训练的预训练的模型，该模型已受过培训但是相关类别。除了发现新颖的课程外，我们还旨在维护模型识别先前看到的基本类别的能力。受到基于彩排的增量学习方法的启发，在本文中，我们提出了一种新颖的方法，以防止通过共同利用基类功能原型和特征级知识蒸馏来忘记对基础类的过去信息。我们还提出了一种自我训练的聚类策略，该策略同时将新颖的类别簇簇，并为基础和新颖类培训共同分类器。这使得我们的方法能够在课堂内设置中运行。我们的实验以三个共同的基准进行，表明我们的方法显着优于最先进的方法。代码可从https://github.com/oatmealliu/class-incd获得

在本文中，我们提出了一种学习内部特征表示模型的新方法，该模型是\ Textit {兼容}与先前学识的。兼容功能可用于直接比较旧和新的学习功能，允许它们随时间互换使用。这消除了在顺序升级表示模型时，可以对视觉搜索系统提取用于在画廊集中的所有先前看到的图像的新功能。在非常大的画廊集和/或实时系统（即面部识别系统，社交网络，终身系统，终身系统，机器人和监测系统）的情况下，提取新功能通常是非常昂贵或不可行的。我们的方法是通过实质性（核心）称为兼容表示，通过鼓励自身定义到学习的表示模型来实现兼容性，而无需依赖以前学习的模型。实用性允许功能在随时间偏移下不改变的统计属性，以便当前学习的功能与旧版本相互操作。我们评估了种植大规模训练数据集中的单一和连续的多模型升级，我们表明我们的方法通过大幅度实现了实现兼容特征来提高现有技术。特别是，通过从Casia-Webface培训和在野外（LFW）中的标记面上评估的培训数据升级十次，我们获得了49 \％的测量倍数达到兼容的平均次数，这是544 \％对先前最先进的相对改善。

深度学习模型在逐步学习新任务时遭受灾难性遗忘。已经提出了增量学习，以保留旧课程的知识，同时学习识别新课程。一种典型的方法是使用一些示例来避免忘记旧知识。在这种情况下，旧类和新课之间的数据失衡是导致模型性能下降的关键问题。由于数据不平衡，已经设计了几种策略来纠正新类别的偏见。但是，他们在很大程度上依赖于新旧阶层之间偏见关系的假设。因此，它们不适合复杂的现实世界应用。在这项研究中，我们提出了一种假设不足的方法，即多粒性重新平衡（MGRB），以解决此问题。重新平衡方法用于减轻数据不平衡的影响；但是，我们从经验上发现，他们将拟合新的课程。为此，我们进一步设计了一个新颖的多晶正式化项，该项使模型还可以考虑除了重新平衡数据之外的类别的相关性。类层次结构首先是通过将语义或视觉上类似类分组来构建的。然后，多粒性正则化将单热标签向量转换为连续的标签分布，这反映了基于构造的类层次结构的目标类别和其他类之间的关系。因此，该模型可以学习类间的关系信息，这有助于增强新旧课程的学习。公共数据集和现实世界中的故障诊断数据集的实验结果验证了所提出的方法的有效性。

在这个不断变化的世界中，必须不断学习新概念的能力。但是，深层神经网络在学习新类别时会遭受灾难性的遗忘。已经提出了许多减轻这种现象的作品，而其中大多数要么属于稳定性困境，要么陷入了过多的计算或储存开销。受到梯度增强算法的启发，以逐渐适应目标模型和上一个合奏模型之间的残差，我们提出了一种新颖的两阶段学习范式寄养物，使该模型能够适应新的类别。具体而言，我们首先动态扩展新模块，以适合原始模型的目标和输出之间的残差。接下来，我们通过有效的蒸馏策略删除冗余参数和特征尺寸，以维护单个骨干模型。我们在不同的设置下验证CIFAR-100和Imagenet-100/1000的方法寄养。实验结果表明，我们的方法实现了最先进的性能。代码可在以下网址获得：https：//github.com/g-u-n/eccv22-foster。

深入学习模型遭受较旧阶段中课程的灾难性遗忘，因为它们在类增量学习设置中新阶段所引入的课程中受过培训。在这项工作中，我们表明灾难性忘记对模型预测的影响随着相同图像的方向的变化而变化，这是一种新的发现。基于此，我们提出了一种新的数据集合方法，该方法结合了图像的不同取向的预测，以帮助模型保留关于先前所见的类别的进一步信息，从而减少忘记模型预测的效果。但是，如果使用传统技术训练，我们无法直接使用数据集合方法。因此，我们还提出了一种新的双重增量学习框架，涉及共同培训网络，其中包括两个增量学习目标，即类渐进式学习目标以及我们提出的数据增量学习目标。在双增量学习框架中，每个图像属于两个类，即图像类（用于类增量学习）和方向类（用于数据增量学习）。在Class-Incremental学习中，每个新阶段都会引入一组新的类，并且模型无法从较旧阶段访问完整的培训数据。在我们提出的数据增量学习中，方向类在所有阶段保持相同，并且在类 - 增量学习中的新阶段引入的数据充当了这些方向类的新培训数据。我们经验证明双增量学习框架对数据集合方法至关重要。我们将拟议的课程逐步增量学习方法应用拟议方法，并经验表明我们的框架显着提高了这些方法的性能。

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。

视觉世界中新对象的不断出现对现实世界部署中当前的深度学习方法构成了巨大的挑战。由于稀有性或成本，新任务学习的挑战通常会加剧新类别的数据。在这里，我们探讨了几乎没有类别学习的重要任务（FSCIL）及其极端数据稀缺条件。理想的FSCIL模型都需要在所有类别上表现良好，无论其显示顺序或数据的匮乏。开放式现实世界条件也需要健壮，并可以轻松地适应始终在现场出现的新任务。在本文中，我们首先重新评估当前的任务设置，并为FSCIL任务提出更全面和实用的设置。然后，受到FSCIL和现代面部识别系统目标的相似性的启发，我们提出了我们的方法 - 增强角损失渐进分类或爱丽丝。在爱丽丝（Alice）中，我们建议使用角度损失损失来获得良好的特征。由于所获得的功能不仅需要紧凑，而且还需要足够多样化以维持未来的增量类别的概括，我们进一步讨论了类增强，数据增强和数据平衡如何影响分类性能。在包括CIFAR100，Miniimagenet和Cub200在内的基准数据集上的实验证明了爱丽丝在最新的FSCIL方法上的性能提高。