Model calibration, which is concerned with how frequently the model predicts correctly, not only plays a vital part in statistical model design, but also has substantial practical applications, such as optimal decision-making in the real world. However, it has been discovered that modern deep neural networks are generally poorly calibrated due to the overestimation (or underestimation) of predictive confidence, which is closely related to overfitting. In this paper, we propose Annealing Double-Head, a simple-to-implement but highly effective architecture for calibrating the DNN during training. To be precise, we construct an additional calibration head-a shallow neural network that typically has one latent layer-on top of the last latent layer in the normal model to map the logits to the aligned confidence. Furthermore, a simple Annealing technique that dynamically scales the logits by calibration head in training procedure is developed to improve its performance. Under both the in-distribution and distributional shift circumstances, we exhaustively evaluate our Annealing Double-Head architecture on multiple pairs of contemporary DNN architectures and vision and speech datasets. We demonstrate that our method achieves state-of-the-art model calibration performance without post-processing while simultaneously providing comparable predictive accuracy in comparison to other recently proposed calibration methods on a range of learning tasks.

Confidence calibration -the problem of predicting probability estimates representative of the true correctness likelihood -is important for classification models in many applications. We discover that modern neural networks, unlike those from a decade ago, are poorly calibrated. Through extensive experiments, we observe that depth, width, weight decay, and Batch Normalization are important factors influencing calibration. We evaluate the performance of various post-processing calibration methods on state-ofthe-art architectures with image and document classification datasets. Our analysis and experiments not only offer insights into neural network learning, but also provide a simple and straightforward recipe for practical settings: on most datasets, temperature scaling -a singleparameter variant of Platt Scaling -is surprisingly effective at calibrating predictions.

尽管深神经网络的占优势性能，但最近的作品表明它们校准不佳，导致过度自信的预测。由于培训期间的跨熵最小化，因此可以通过过度化来加剧错误烫伤，因为它促进了预测的Softmax概率来匹配单热标签分配。这产生了正确的类别的Pre-SoftMax激活，该类别明显大于剩余的激活。来自文献的最近证据表明，损失函数嵌入隐含或明确最大化的预测熵会产生最先进的校准性能。我们提供了当前最先进的校准损耗的统一约束优化视角。具体地，这些损失可以被视为在Logit距离上施加平等约束的线性惩罚（或拉格朗日）的近似值。这指出了这种潜在的平等约束的一个重要限制，其随后的梯度不断推动非信息解决方案，这可能会阻止在基于梯度的优化期间模型的辨别性能和校准之间的最佳妥协。在我们的观察之后，我们提出了一种基于不平等约束的简单灵活的泛化，这在Logit距离上强加了可控裕度。关于各种图像分类，语义分割和NLP基准的综合实验表明，我们的方法在网络校准方面对这些任务设置了新的最先进的结果，而不会影响辨别性能。代码可在https://github.com/by-liu/mbls上获得。

神经网络校准是深度学习的重要任务，以确保模型预测的信心与真正的正确性可能性之间的一致性。在本文中，我们提出了一种称为Neural夹紧的新的后处理校准方法，该方法通过可学习的通用输入扰动和输出温度扩展参数在预训练的分类器上采用简单的联合输入输出转换。此外，我们提供了理论上的解释，说明为什么神经夹具比温度缩放更好。在CIFAR-100和Imagenet图像识别数据集以及各种深神经网络模型上进行了评估，我们的经验结果表明，神经夹具明显优于最先进的后处理校准方法。

Recent studies have revealed that, beyond conventional accuracy, calibration should also be considered for training modern deep neural networks. To address miscalibration during learning, some methods have explored different penalty functions as part of the learning objective, alongside a standard classification loss, with a hyper-parameter controlling the relative contribution of each term. Nevertheless, these methods share two major drawbacks: 1) the scalar balancing weight is the same for all classes, hindering the ability to address different intrinsic difficulties or imbalance among classes; and 2) the balancing weight is usually fixed without an adaptive strategy, which may prevent from reaching the best compromise between accuracy and calibration, and requires hyper-parameter search for each application. We propose Class Adaptive Label Smoothing (CALS) for calibrating deep networks, which allows to learn class-wise multipliers during training, yielding a powerful alternative to common label smoothing penalties. Our method builds on a general Augmented Lagrangian approach, a well-established technique in constrained optimization, but we introduce several modifications to tailor it for large-scale, class-adaptive training. Comprehensive evaluation and multiple comparisons on a variety of benchmarks, including standard and long-tailed image classification, semantic segmentation, and text classification, demonstrate the superiority of the proposed method. The code is available at https://github.com/by-liu/CALS.

在本文中，我们研究了现代神经网络的事后校准，这个问题近年来引起了很多关注。已经为任务提出了许多不同复杂性的校准方法，但是关于这些任务的表达方式尚无共识。我们专注于置信度缩放的任务，特别是在概括温度缩放的事后方法上，我们将其称为自适应温度缩放家族。我们分析了改善校准并提出可解释方法的表达功能。我们表明，当有大量数据复杂模型（例如神经网络）产生更好的性能时，但是当数据量受到限制时，很容易失败，这是某些事后校准应用（例如医学诊断）的常见情况。我们研究表达方法在理想条件和设计更简单的方法下学习但对这些表现良好的功能具有强烈的感应偏见的功能。具体而言，我们提出了基于熵的温度缩放，这是一种简单的方法，可根据其熵缩放预测的置信度。结果表明，与其他方法相比，我们的方法可获得最先进的性能，并且与复杂模型不同，它对数据稀缺是可靠的。此外，我们提出的模型可以更深入地解释校准过程。

现在众所周知，神经网络对其预测的信心很高，导致校准不良。弥补这一点的最常见的事后方法是执行温度缩放，这可以通过将逻辑缩放为固定值来调整任何输入的预测的信心。尽管这种方法通常会改善整个测试数据集中的平均校准，但无论给定输入的分类是否正确还是不正确，这种改进通常会降低预测的个人信心。有了这种见解，我们将方法基于这样的观察结果，即不同的样品通过不同的量导致校准误差，有些人需要提高其信心，而另一些则需要减少它。因此，对于每个输入，我们建议预测不同的温度值，从而使我们能够调整较细性的置信度和准确性之间的不匹配。此外，我们观察到了OOD检测结果的改善，还可以提取数据点的硬度概念。我们的方法是在事后应用的，因此使用很少的计算时间和可忽略不计的记忆足迹，并应用于现成的预训练的分类器。我们使用CIFAR10/100和TINY-IMAGENET数据集对RESNET50和WIDERESNET28-10架构进行测试，这表明在整个测试集中产生每数据点温度也有益于预期的校准误差。代码可在以下网址获得：https：//github.com/thwjoy/adats。

深度神经网络具有令人印象深刻的性能，但是他们无法可靠地估计其预测信心，从而限制了其在高风险领域中的适用性。我们表明，应用多标签的一VS损失揭示了分类的歧义并降低了模型的过度自信。引入的Slova（单标签One-Vs-All）模型重新定义了单个标签情况的典型单VS-ALL预测概率，其中只有一个类是正确的答案。仅当单个类具有很高的概率并且其他概率可忽略不计时，提议的分类器才有信心。与典型的SoftMax函数不同，如果所有其他类的概率都很小，Slova自然会检测到分布的样本。该模型还通过指数校准进行了微调，这使我们能够与模型精度准确地对齐置信分数。我们在三个任务上验证我们的方法。首先，我们证明了斯洛伐克与最先进的分布校准具有竞争力。其次，在数据集偏移下，斯洛伐克的性能很强。最后，我们的方法在检测到分布样品的检测方面表现出色。因此，斯洛伐克是一种工具，可以在需要不确定性建模的各种应用中使用。

最佳决策要求分类器产生与其经验准确性一致的不确定性估计。然而，深度神经网络通常在他们的预测中受到影响或过度自信。因此，已经开发了方法，以改善培训和后HOC期间的预测性不确定性的校准。在这项工作中，我们提出了可分解的损失，以改善基于频流校准误差估计底层的钻孔操作的软（连续）版本的校准。当纳入训练时，这些软校准损耗在多个数据集中实现最先进的单一模型ECE，精度低于1％的数量。例如，我们观察到ECE的82％（相对于HOC后射出ECE 70％），以换取相对于CIFAR-100上的交叉熵基线的准确性0.7％的相对降低。在培训后结合时，基于软合成的校准误差目标会改善温度缩放，一种流行的重新校准方法。总体而言，跨损失和数据集的实验表明，使用校准敏感程序在数据集移位下产生更好的不确定性估计，而不是使用跨熵损失和后HOC重新校准方法的标准做法。

Modern machine learning methods including deep learning have achieved great success in predictive accuracy for supervised learning tasks, but may still fall short in giving useful estimates of their predictive uncertainty. Quantifying uncertainty is especially critical in real-world settings, which often involve input distributions that are shifted from the training distribution due to a variety of factors including sample bias and non-stationarity. In such settings, well calibrated uncertainty estimates convey information about when a model's output should (or should not) be trusted. Many probabilistic deep learning methods, including Bayesian-and non-Bayesian methods, have been proposed in the literature for quantifying predictive uncertainty, but to our knowledge there has not previously been a rigorous largescale empirical comparison of these methods under dataset shift. We present a largescale benchmark of existing state-of-the-art methods on classification problems and investigate the effect of dataset shift on accuracy and calibration. We find that traditional post-hoc calibration does indeed fall short, as do several other previous methods. However, some methods that marginalize over models give surprisingly strong results across a broad spectrum of tasks.

Deep neural networks (DNN) are prone to miscalibrated predictions, often exhibiting a mismatch between the predicted output and the associated confidence scores. Contemporary model calibration techniques mitigate the problem of overconfident predictions by pushing down the confidence of the winning class while increasing the confidence of the remaining classes across all test samples. However, from a deployment perspective, an ideal model is desired to (i) generate well-calibrated predictions for high-confidence samples with predicted probability say >0.95, and (ii) generate a higher proportion of legitimate high-confidence samples. To this end, we propose a novel regularization technique that can be used with classification losses, leading to state-of-the-art calibrated predictions at test time; From a deployment standpoint in safety-critical applications, only high-confidence samples from a well-calibrated model are of interest, as the remaining samples have to undergo manual inspection. Predictive confidence reduction of these potentially ``high-confidence samples'' is a downside of existing calibration approaches. We mitigate this by proposing a dynamic train-time data pruning strategy that prunes low-confidence samples every few epochs, providing an increase in "confident yet calibrated samples". We demonstrate state-of-the-art calibration performance across image classification benchmarks, reducing training time without much compromise in accuracy. We provide insights into why our dynamic pruning strategy that prunes low-confidence training samples leads to an increase in high-confidence samples at test time.

我们表明，著名的混音的有效性[Zhang等，2018]，如果而不是将其用作唯一的学习目标，就可以进一步改善它，而是将其用作标准跨侧面损失的附加规则器。这种简单的变化不仅提供了太大的准确性，而且在大多数情况下，在各种形式的协变量转移和分布外检测实验下，在大多数情况下，混合量的预测不确定性估计质量都显着提高了。实际上，我们观察到混合物在检测出分布样本时可能会产生大量退化的性能，因为我们在经验上表现出来，因为它倾向于学习在整个过程中表现出高渗透率的模型。很难区分分布样本与近分离样本。为了显示我们的方法的功效（RegMixup），我们在视觉数据集（Imagenet＆Cifar-10/100）上提供了详尽的分析和实验，并将其与最新方法进行比较，以进行可靠的不确定性估计。

检测到分布输入对于在现实世界中安全部署机器学习模型至关重要。然而，已知神经网络遭受过度自信的问题，在该问题中，它们对分布和分布的输入的信心异常高。在这项工作中，我们表明，可以通过在训练中实施恒定的向量规范来通过logit归一化（logitnorm）（logitnorm）来缓解此问题。我们的方法是通过分析的激励，即logit的规范在训练过程中不断增加，从而导致过度自信的产出。因此，LogitNorm背后的关键思想是将网络优化期间输出规范的影响解散。通过LogitNorm培训，神经网络在分布数据和分布数据之间产生高度可区分的置信度得分。广泛的实验证明了LogitNorm的优势，在公共基准上，平均FPR95最高为42.30％。

We propose SWA-Gaussian (SWAG), a simple, scalable, and general purpose approach for uncertainty representation and calibration in deep learning. Stochastic Weight Averaging (SWA), which computes the first moment of stochastic gradient descent (SGD) iterates with a modified learning rate schedule, has recently been shown to improve generalization in deep learning. With SWAG, we fit a Gaussian using the SWA solution as the first moment and a low rank plus diagonal covariance also derived from the SGD iterates, forming an approximate posterior distribution over neural network weights; we then sample from this Gaussian distribution to perform Bayesian model averaging. We empirically find that SWAG approximates the shape of the true posterior, in accordance with results describing the stationary distribution of SGD iterates. Moreover, we demonstrate that SWAG performs well on a wide variety of tasks, including out of sample detection, calibration, and transfer learning, in comparison to many popular alternatives including MC dropout, KFAC Laplace, SGLD, and temperature scaling.

Deep neural network (DNN) classifiers are often overconfident, producing miscalibrated class probabilities. Most existing calibration methods either lack theoretical guarantees for producing calibrated outputs or reduce the classification accuracy in the process. This paper proposes a new Kernel-based calibration method called KCal. Unlike other calibration procedures, KCal does not operate directly on the logits or softmax outputs of the DNN. Instead, it uses the penultimate-layer latent embedding to train a metric space in a supervised manner. In effect, KCal amounts to a supervised dimensionality reduction of the neural network embedding, and generates a prediction using kernel density estimation on a holdout calibration set. We first analyze KCal theoretically, showing that it enjoys a provable asymptotic calibration guarantee. Then, through extensive experiments, we confirm that KCal consistently outperforms existing calibration methods in terms of both the classification accuracy and the (confidence and class-wise) calibration error.

神经网络缺乏对抗性鲁棒性，即，它们容易受到对抗的例子，通过对输入的小扰动导致错误的预测。此外，当模型给出错误的预测时，信任被破坏，即，预测的概率不是我们应该相信我们模型的良好指标。在本文中，我们研究了对抗性鲁棒性和校准之间的联系，发现模型对小扰动敏感的输入（很容易攻击）更有可能具有较差的预测。基于这种洞察力，我们通过解决这些对抗的缺陷输入来研究校准。为此，我们提出了基于对抗基于对抗的自适应标签平滑（AR-AD），其通过适应性软化标签，通过适应性软化标签来整合对抗性鲁棒性和校准到训练中的相关性，这是基于对敌人可以攻击的容易攻击。我们发现我们的方法，考虑了分销数据的对抗性稳健性，即使在分布班次下也能够更好地校准模型。此外，还可以应用于集合模型，以进一步提高模型校准。

许多现实世界的识别问题都有不平衡或长尾标签的分布。这些分布使表示形式学习更具挑战性，因为对尾巴类别的概括有限。如果测试分布与训练分布有所不同，例如统一与长尾，需要解决分配转移的问题。为此，最近的作品通过贝叶斯定理的启发，使用边缘修改扩展了SoftMax跨凝结。在本文中，我们通过专家的平衡产品（Balpoe）概括了几种方法，该方法结合了一个具有不同测试时间目标分布的模型家庭，以解决数据中的不平衡。拟议的专家在一个阶段进行培训，无论是共同还是独立的，并无缝融合到Balpoe中。我们表明，Balpoe是Fisher的一致性，可以最大程度地减少均衡误差并执行广泛的实验以验证我们的方法的有效性。最后，我们研究了在这种情况下混合的效果，发现正则化是学习校准专家的关键要素。我们的实验表明，正则化的BALPOE在测试准确性和校准指标上的表现非常出色，从而导致CIFAR-100-LT，Imagenet-LT和Inaturalist-2018数据集的最新结果。该代码将在纸质接受后公开提供。

我们解决了不确定性校准的问题，并引入了一种新型的校准方法，即参数化温度缩放（PTS）。标准的深神经网络通常会产生未校准的预测，可以使用事后校准方法将其转化为校准的置信得分。在这项贡献中，我们证明了准确保存最先进的事后校准器的性能受其内在表达能力的限制。我们通过计算通过神经网络参数为参数的预测温度来概括温度缩放。我们通过广泛的实验表明，我们的新型准确性保护方法始终优于大量模型体系结构，数据集和指标的现有算法。

深度神经网络易于对异常值过度自信的预测。贝叶斯神经网络和深度融合都已显示在某种程度上减轻了这个问题。在这项工作中，我们的目标是通过提议预测由高斯混合模型的后续的高斯混合模型来结合这两种方法的益处，该高斯混合模型包括独立培训的深神经网络的LAPPALL近似的加权和。该方法可以与任何一组预先训练的网络一起使用，并且与常规合并相比，只需要小的计算和内存开销。理论上我们验证了我们的方法从训练数据中的培训数据和虚拟化的基本线上的标准不确定量级基准测试中的“远离”的过度控制。

Accurate uncertainty quantification is a major challenge in deep learning, as neural networks can make overconfident errors and assign high confidence predictions to out-of-distribution (OOD) inputs. The most popular approaches to estimate predictive uncertainty in deep learning are methods that combine predictions from multiple neural networks, such as Bayesian neural networks (BNNs) and deep ensembles. However their practicality in real-time, industrial-scale applications are limited due to the high memory and computational cost. Furthermore, ensembles and BNNs do not necessarily fix all the issues with the underlying member networks. In this work, we study principled approaches to improve uncertainty property of a single network, based on a single, deterministic representation. By formalizing the uncertainty quantification as a minimax learning problem, we first identify distance awareness, i.e., the model's ability to quantify the distance of a testing example from the training data, as a necessary condition for a DNN to achieve high-quality (i.e., minimax optimal) uncertainty estimation. We then propose Spectral-normalized Neural Gaussian Process (SNGP), a simple method that improves the distance-awareness ability of modern DNNs with two simple changes: (1) applying spectral normalization to hidden weights to enforce bi-Lipschitz smoothness in representations and (2) replacing the last output layer with a Gaussian process layer. On a suite of vision and language understanding benchmarks, SNGP outperforms other single-model approaches in prediction, calibration and out-of-domain detection. Furthermore, SNGP provides complementary benefits to popular techniques such as deep ensembles and data augmentation, making it a simple and scalable building block for probabilistic deep learning. Code is open-sourced at https://github.com/google/uncertainty-baselines