深度神经网络对物体检测达到了高精度，但它们的成功铰链大量标记数据。为了减少标签依赖性，已经提出了各种主动学习策略，通常基于探测器的置信度。但是，这些方法偏向于高性能类，并且可以导致获取的数据集不是测试集数据的代表不好。在这项工作中，我们提出了一个统一的主动学习框架，这考虑了探测器的不确定性和鲁棒性，确保网络在所有类中表现良好。此外，我们的方法利用自动标记来抑制潜在的分布漂移，同时提高模型的性能。 Pascal VOC07 ​​+ 12和MS-Coco的实验表明，我们的方法始终如一地优于各种活跃的学习方法，在地图中产生高达7.7％，或降低标记成本的82％。代码将在接受纸张时发布。

接受注释较弱的对象探测器是全面监督者的负担得起的替代方案。但是，它们之间仍然存在显着的性能差距。我们建议通过微调预先训练的弱监督检测器来缩小这一差距，并使用``Box-In-box''（bib'（bib）自动从训练集中自动选择了一些完全注销的样品，这是一种新颖的活跃学习专门针对弱势监督探测器的据可查的失败模式而设计的策略。 VOC07和可可基准的实验表明，围嘴表现优于其他活跃的学习技术，并显着改善了基本的弱监督探测器的性能，而每个类别仅几个完全宣布的图像。围嘴达到了完全监督的快速RCNN的97％，在VOC07上仅10％的全已通量图像。在可可（COCO）上，平均每类使用10张全面通量的图像，或同等的训练集的1％，还减少了弱监督检测器和完全监督的快速RCN之间的性能差距（In AP）以上超过70％ ，在性能和数据效率之间表现出良好的权衡。我们的代码可在https://github.com/huyvvo/bib上公开获取。

As an important data selection schema, active learning emerges as the essential component when iterating an Artificial Intelligence (AI) model. It becomes even more critical given the dominance of deep neural network based models, which are composed of a large number of parameters and data hungry, in application. Despite its indispensable role for developing AI models, research on active learning is not as intensive as other research directions. In this paper, we present a review of active learning through deep active learning approaches from the following perspectives: 1) technical advancements in active learning, 2) applications of active learning in computer vision, 3) industrial systems leveraging or with potential to leverage active learning for data iteration, 4) current limitations and future research directions. We expect this paper to clarify the significance of active learning in a modern AI model manufacturing process and to bring additional research attention to active learning. By addressing data automation challenges and coping with automated machine learning systems, active learning will facilitate democratization of AI technologies by boosting model production at scale.

The performance of deep neural networks improves with more annotated data. The problem is that the budget for annotation is limited. One solution to this is active learning, where a model asks human to annotate data that it perceived as uncertain. A variety of recent methods have been proposed to apply active learning to deep networks but most of them are either designed specific for their target tasks or computationally inefficient for large networks. In this paper, we propose a novel active learning method that is simple but task-agnostic, and works efficiently with the deep networks. We attach a small parametric module, named "loss prediction module," to a target network, and learn it to predict target losses of unlabeled inputs. Then, this module can suggest data that the target model is likely to produce a wrong prediction. This method is task-agnostic as networks are learned from a single loss regardless of target tasks. We rigorously validate our method through image classification, object detection, and human pose estimation, with the recent network architectures. The results demonstrate that our method consistently outperforms the previous methods over the tasks.

Deep learning has emerged as an effective solution for solving the task of object detection in images but at the cost of requiring large labeled datasets. To mitigate this cost, semi-supervised object detection methods, which consist in leveraging abundant unlabeled data, have been proposed and have already shown impressive results. However, most of these methods require linking a pseudo-label to a ground-truth object by thresholding. In previous works, this threshold value is usually determined empirically, which is time consuming, and only done for a single data distribution. When the domain, and thus the data distribution, changes, a new and costly parameter search is necessary. In this work, we introduce our method Adaptive Self-Training for Object Detection (ASTOD), which is a simple yet effective teacher-student method. ASTOD determines without cost a threshold value based directly on the ground value of the score histogram. To improve the quality of the teacher predictions, we also propose a novel pseudo-labeling procedure. We use different views of the unlabeled images during the pseudo-labeling step to reduce the number of missed predictions and thus obtain better candidate labels. Our teacher and our student are trained separately, and our method can be used in an iterative fashion by replacing the teacher by the student. On the MS-COCO dataset, our method consistently performs favorably against state-of-the-art methods that do not require a threshold parameter, and shows competitive results with methods that require a parameter sweep search. Additional experiments with respect to a supervised baseline on the DIOR dataset containing satellite images lead to similar conclusions, and prove that it is possible to adapt the score threshold automatically in self-training, regardless of the data distribution.

Semi-supervised object detection (SSOD) aims to boost detection performance by leveraging extra unlabeled data. The teacher-student framework has been shown to be promising for SSOD, in which a teacher network generates pseudo-labels for unlabeled data to assist the training of a student network. Since the pseudo-labels are noisy, filtering the pseudo-labels is crucial to exploit the potential of such framework. Unlike existing suboptimal methods, we propose a two-step pseudo-label filtering for the classification and regression heads in a teacher-student framework. For the classification head, OCL (Object-wise Contrastive Learning) regularizes the object representation learning that utilizes unlabeled data to improve pseudo-label filtering by enhancing the discriminativeness of the classification score. This is designed to pull together objects in the same class and push away objects from different classes. For the regression head, we further propose RUPL (Regression-Uncertainty-guided Pseudo-Labeling) to learn the aleatoric uncertainty of object localization for label filtering. By jointly filtering the pseudo-labels for the classification and regression heads, the student network receives better guidance from the teacher network for object detection task. Experimental results on Pascal VOC and MS-COCO datasets demonstrate the superiority of our proposed method with competitive performance compared to existing methods.

迄今为止，最强大的半监督对象检测器（SS-OD）基于伪盒，该盒子需要一系列带有微调超参数的后处理。在这项工作中，我们建议用稀疏的伪盒子以伪造的伪标签形式取代稀疏的伪盒。与伪盒相比，我们的密集伪标签（DPL）不涉及任何后处理方法，因此保留了更丰富的信息。我们还引入了一种区域选择技术，以突出关键信息，同时抑制密集标签所携带的噪声。我们将利用DPL作为密集老师的拟议的SS-OD算法命名。在可可和VOC上，密集的老师在各种环境下与基于伪盒的方法相比表现出卓越的表现。

尽管半监督学习（SSL）的最新研究已经在单标签分类问题上取得了强劲的表现，但同样重要但毫无疑问的问题是如何利用多标签分类任务中未标记数据的优势。为了将SSL的成功扩展到多标签分类，我们首先使用说明性示例进行分析，以获得有关多标签分类中存在的额外挑战的一些直觉。基于分析，我们提出了一个基于百分比的阈值调整方案的百分位摩擦，以动态地改变训练期间每个类别的正和负伪标签的得分阈值，以及动态的未标记失误权重，从而进一步降低了从早期未标记的预测。与最近的SSL方法相比，在不丧失简单性的情况下，我们在Pascal VOC2007和MS-Coco数据集上实现了强劲的性能。

随着半监督对象检测（SS-OD）技术的最新开发，可以使用有限的标记数据和丰富的未标记数据来改进对象检测器。但是，仍然有两个挑战未解决：（1）在无锚点检测器上没有先前的SS-OD作品，并且（2）当伪标记的边界框回归时，先前的工作是无效的。在本文中，我们提出了无偏见的教师V2，其中显示了SS-OD方法对无锚定检测器的概括，并引入了无监督回归损失的侦听机制。具体而言，我们首先提出了一项研究，研究了现有的SS-OD方法在无锚固探测器上的有效性，并发现在半监督的设置下它们的性能改善要较低。我们还观察到，在无锚点检测器中使用的中心度和基于本地化的标签的盒子选择不能在半监视的设置下正常工作。另一方面，我们的聆听机制明确地阻止了在边界框回归训练中误导伪标记。我们特别开发了一种基于教师和学生的相对不确定性的新型伪标记的选择机制。这个想法有助于半监督环境中回归分支的有利改善。我们的方法适用于无锚固方法和基于锚的方法，它始终如一地对VOC，可可标准和可可添加的最新方法表现出色。

研究表明，当训练数据缺少注释时，对象检测器的性能下降，即稀疏注释数据。当代方法专注于缺少地面实话注释的代理，无论是伪标签的形式还是通过在训练期间重新称重梯度。在这项工作中，我们重新审视了稀疏注释物体检测的制定。我们观察到稀疏注释的物体检测可以被认为是区域级的半监督对象检测问题。在此洞察力上，我们提出了一种基于区域的半监督算法，它自动识别包含未标记的前景对象的区域。我们的算法然后以不同的方式处理标记和未标记的前景区域，在半监督方法中进行常见做法。为了评估所提出的方法的有效性，我们对普斯卡尔库尔和可可数据集的稀疏注释方法常用的五种分裂进行详尽的实验，并实现最先进的性能。除此之外，我们还表明，我们的方法在标准半监督设置上实现了竞争性能，证明了我们的方法的实力和广泛适用性。

半弱监督和监督的学习最近在对象检测文献中引起了很大的关注，因为它们可以减轻成功训练深度学习模型所需的注释成本。半监督学习的最先进方法依赖于使用多阶段过程训练的学生老师模型，并大量数据增强。为弱监督的设置开发了自定义网络，因此很难适应不同的检测器。在本文中，引入了一种弱半监督的训练方法，以减少这些训练挑战，但通过仅利用一小部分全标记的图像，并在弱标记图像中提供信息来实现最先进的性能。特别是，我们基于通用抽样的学习策略以在线方式产生伪基真实（GT）边界框注释，消除了对多阶段培训的需求和学生教师网络配置。这些伪GT框是根据通过得分传播过程累积的对象建议的分类得分从弱标记的图像中采样的。 PASCAL VOC数据集的经验结果表明，使用VOC 2007作为完全标记的拟议方法可提高性能5.0％，而VOC 2012作为弱标记数据。同样，有了5-10％的完全注释的图像，我们观察到MAP中的10％以上的改善，表明对图像级注释的适度投资可以大大改善检测性能。

Recent aerial object detection models rely on a large amount of labeled training data, which requires unaffordable manual labeling costs in large aerial scenes with dense objects. Active learning is effective in reducing the data labeling cost by selectively querying the informative and representative unlabelled samples. However, existing active learning methods are mainly with class-balanced setting and image-based querying for generic object detection tasks, which are less applicable to aerial object detection scenario due to the long-tailed class distribution and dense small objects in aerial scenes. In this paper, we propose a novel active learning method for cost-effective aerial object detection. Specifically, both object-level and image-level informativeness are considered in the object selection to refrain from redundant and myopic querying. Besides, an easy-to-use class-balancing criterion is incorporated to favor the minority objects to alleviate the long-tailed class distribution problem in model training. To fully utilize the queried information, we further devise a training loss to mine the latent knowledge in the undiscovered image regions. Extensive experiments are conducted on the DOTA-v1.0 and DOTA-v2.0 benchmarks to validate the effectiveness of the proposed method. The results show that it can save more than 75% of the labeling cost to reach the same performance compared to the baselines and state-of-the-art active object detection methods. Code is available at https://github.com/ZJW700/MUS-CDB

利用伪标签（例如，类别和边界框）由教师探测器产生的未注释的对象，已经为半监督对象检测（SSOD）的最新进展提供了很多进展。但是，由于稀缺注释引起的教师探测器的概括能力有限，因此产生的伪标签通常偏离地面真理，尤其是那些具有相对较低分类信心的人，从而限制了SSOD的概括性能。为了减轻此问题，我们为SSOD提出了一个双伪标签抛光框架。我们没有直接利用教师探测器生成的伪标签，而是首次尝试使用双抛光学习来减少它们偏离地面真相的偏差，其中两个不同结构化的抛光网络是精心开发和培训的分别在给定注释对象上的类别和边界框的真相。通过这样做，两个抛光网络都可以通过基于最初产生的伪标签充分利用其上下文知识来推断未注释的对象的更准确的伪标签，从而提高了SSOD的概括性能。此外，可以将这种方案无缝地插入现有的SSOD框架中，以进行端到端学习。此外，我们建议将抛光的伪类别和未注释的对象的边界框，用于单独的类别分类和SSOD中的边界框回归，这使得在模型训练过程中可以引入更多未经许可的对象，从而进一步提高了性能。 Pascal VOC和MS Coco基准测试的实验证明了该方法比现有最新基准的优越性。

Object detection requires substantial labeling effort for learning robust models. Active learning can reduce this effort by intelligently selecting relevant examples to be annotated. However, selecting these examples properly without introducing a sampling bias with a negative impact on the generalization performance is not straightforward and most active learning techniques can not hold their promises on real-world benchmarks. In our evaluation paper, we focus on active learning techniques without a computational overhead besides inference, something we refer to as zero-cost active learning. In particular, we show that a key ingredient is not only the score on a bounding box level but also the technique used for aggregating the scores for ranking images. We outline our experimental setup and also discuss practical considerations when using active learning for object detection.

Active learning as a paradigm in deep learning is especially important in applications involving intricate perception tasks such as object detection where labels are difficult and expensive to acquire. Development of active learning methods in such fields is highly computationally expensive and time consuming which obstructs the progression of research and leads to a lack of comparability between methods. In this work, we propose and investigate a sandbox setup for rapid development and transparent evaluation of active learning in deep object detection. Our experiments with commonly used configurations of datasets and detection architectures found in the literature show that results obtained in our sandbox environment are representative of results on standard configurations. The total compute time to obtain results and assess the learning behavior can thereby be reduced by factors of up to 14 when comparing with Pascal VOC and up to 32 when comparing with BDD100k. This allows for testing and evaluating data acquisition and labeling strategies in under half a day and contributes to the transparency and development speed in the field of active learning for object detection.

在这项研究中，我们深入研究了半监督对象检测〜（SSOD）所面临的独特挑战。我们观察到当前的探测器通常遭受3个不一致问题。 1）分配不一致，传统的分配策略对标记噪声很敏感。 2）子任务不一致，其中分类和回归预测在同一特征点未对准。 3）时间不一致，伪Bbox在不同的训练步骤中差异很大。这些问题导致学生网络的优化目标不一致，从而恶化了性能并减慢模型收敛性。因此，我们提出了一个系统的解决方案，称为一致的老师，以补救上述挑战。首先，自适应锚分配代替了基于静态的策略，该策略使学生网络能够抵抗嘈杂的psudo bbox。然后，我们通过设计功能比对模块来校准子任务预测。最后，我们采用高斯混合模型（GMM）来动态调整伪盒阈值。一致的老师在各种SSOD评估上提供了新的强大基线。只有10％的带注释的MS-Coco数据，它可以使用Resnet-50骨干实现40.0 MAP，该数据仅使用伪标签，超过了4个地图。当对完全注释的MS-Coco进行其他未标记的数据进行培训时，性能将进一步增加到49.1 MAP。我们的代码将很快开源。

最近，许多半监督的对象检测（SSOD）方法采用教师学生框架并取得了最新的结果。但是，教师网络与学生网络紧密相结合，因为教师是学生的指数移动平均值（EMA），这会导致表现瓶颈。为了解决耦合问题，我们为SSOD提出了一个周期自我训练（CST）框架，该框架由两个老师T1和T2，两个学生S1和S2组成。基于这些网络，构建了一个周期自我训练机制​​，即S1 $ {\ rightarrow} $ t1 $ {\ rightArow} $ s2 $ {\ rightArrow} $ t2 $ {\ rightArrow} $ s1。对于S $ {\ Rightarrow} $ T，我们还利用学生的EMA权重来更新老师。对于t $ {\ rightarrow} $ s，而不是直接为其学生S1（S2）提供监督，而是老师T1（T2）为学生S2（S1）生成伪标记，从而松散耦合效果。此外，由于EMA的财产，老师最有可能积累学生的偏见，并使错误变得不可逆转。为了减轻问题，我们还提出了分配一致性重新加权策略，在该策略中，根据教师T1和T2的分配一致性，将伪标记重新加权。通过该策略，可以使用嘈杂的伪标签对两个学生S2和S1进行训练，以避免确认偏见。广泛的实验证明了CST的优势，通过将AP比基线优于最先进的方法提高了2.1％的绝对AP改进，并具有稀缺的标记数据，而胜过了2.1％的绝对AP。

半监督学习旨在利用大量未标记的数据进行性能提升。现有工作主要关注图像分类。在本文中，我们深入了解对象检测的半监督学习，其中标记的数据更加劳动密集。目前的方法是由伪标签产生的嘈杂区域分散注意力。为了打击嘈杂的标签，我们通过量化区域不确定性提出抗噪声的半监督学习。我们首先调查与伪标签相关的不同形式的噪声带来的不利影响。然后，我们建议通过识别不同强度的区域的抗性特性来量化区域的不确定性。通过导入该地区不确定性量化和促进多跳概率分布输出，我们将不确定性引入训练和进一步实现抗噪声学习。 Pascal VOC和MS COCO两者的实验证明了我们的方法的特殊表现。

半监督学习（SSL）的最新最新方法将一致性正则化与基于置信的伪标记结合在一起。为了获得高质量的伪标签，通常采用高置信度阈值。但是，已经表明，对于远离训练数据的样本，深网的基于软磁性的置信度得分可能很高，因此，即使是高信心不明的样品，伪标签也可能仍然不可靠。在这项工作中，我们提出了伪标记的新观点：而不是依靠模型信心，而是衡量未标记的样本是否可能是“分布”；即，接近当前的培训数据。为了对未标记的样本进行分类是“分布”还是“分发”，我们采用了分布外检测文献中的能量评分。随着培训的进行进展，更不标记的样品成为分配并有助于培训，标记和伪标记的数据可以更好地近似于真正的分布以改善模型。实验表明，我们的基于能量的伪标记方法，尽管从概念上讲简单，但在不平衡的SSL基准测试方面显着优于基于置信的方法，并在类平衡的数据上实现了竞争性能。例如，当不平衡比率高于50时，它会在CIFAR10-LT上产生4-6％的绝对准确性提高。当与最新的长尾SSL方法结合使用时，可以实现进一步的改进。

半监督学习（SSL）是一个有效的框架，可以使用标记和未标记的数据训练模型，但是当缺乏足够的标记样品时，可能会产生模棱两可和不可区分的表示。有了人类的循环学习，积极的学习可以迭代地选择无标记的样品进行标签和培训，以提高SSL框架的性能。但是，大多数现有的活跃学习方法都取决于预先训练的功能，这不适合端到端学习。为了解决SSL的缺点，在本文中，我们提出了一种新颖的端到端表示方法，即ActiveMatch，它将SSL与对比度学习和积极学习结合在一起，以充分利用有限的标签。从少量的标记数据开始，无监督的对比度学习作为热身学习，然后将ActiveMatch结合在一起，将SSL和监督对比度学习结合在一起，并积极选择在培训期间标记的最具代表性的样本，从而更好地表示分类。与MixMatch和FixMatch具有相同数量的标记数据相比，我们表明ActiveMatch实现了最先进的性能，CIFAR-10的精度为89.24％，具有100个收集的标签，而92.20％的精度为92.20％，有200个收集的标签。