blocks|key|1210974|text|只在训练时。数据增强用于增加训练集的大小，并获得更多不同的图像。从技术上讲，您可以使用测试集上的数据增强来查看模型在这些图像上的行为，但通常，人们不会这样做。|type|unstyled|depth|inlineStyleRanges|entityRanges|data|1210975|entityMap^0|0^^$0|@$1|2|3|4|5|6|7|D|8|@]|9|@]|A|$]]|$1|B|3|-4|5|6|7|E|8|@]|9|@]|A|$]]]|C|$]]

Only on training. Data augmentation is used to increase the size of the training set and to get more different images.
Technically, you could use data augmentation on the test set to see how the model behaves on such images, but usually, people don't do it.

blocks|key|608138|text|这个答案在stats.SE上为在验证/测试集上应用作物提供了理由，以便使该输入与网络所受培训集中的输入相似。|type|unstyled|depth|inlineStyleRanges|entityRanges|offset|length|data|608139|entityMap|0|LINK|mutability|MUTABLE|url|https://stats.stackexchange.com/a/320967/162527^0|0|E|0|0^^$0|@$1|2|3|4|5|6|7|L|8|@]|9|@$A|M|B|N|1|O]]|C|$]]|$1|D|3|-4|5|6|7|P|8|@]|9|@]|C|$]]]|E|$F|$5|G|H|I|C|$J|K]]]]

<a href="https://stats.stackexchange.com/a/320967/162527">This answer on stats.SE</a> makes the case for applying crops on the validation / test sets so as to make that input similar the the input in the training set that the network was trained on.

blocks|key|1210990|text|数据增强只在训练集上进行，因为它帮助模型变得更一般化和更健壮。所以增加测试集是没有意义的。|type|unstyled|depth|inlineStyleRanges|entityRanges|data|1210991|entityMap^0|0^^$0|@$1|2|3|4|5|6|7|D|8|@]|9|@]|A|$]]|$1|B|3|-4|5|6|7|E|8|@]|9|@]|A|$]]]|C|$]]

Data augmentation is done only on training set as it helps the model become more generalize and robust. So there's no point of augmenting the test set.

blocks|key|1211015|text|只在训练组做。当然，也要确保增强不会使标签错误(例如，当6和9旋转大约180°时)。|type|unstyled|depth|inlineStyleRanges|entityRanges|data|1211016|首先，我们使用培训和测试集的原因是，我们想要估计我们的系统在现实中会遇到的错误。因此，测试集的数据应该尽可能接近真实数据。|1211017|如果在测试集中执行此操作，可能会出现引入错误的问题。例如，假设您想要识别数字，然后通过旋转进行增强。然后，6可能看起来像9。但并非所有的例子都那么容易。最好是拯救而不是抱歉。|offset|length|style|CODE|1211018|entityMap^0|0|0|1H|1|1O|1|0^^$0|@$1|2|3|4|5|6|7|L|8|@]|9|@]|A|$]]|$1|B|3|C|5|6|7|M|8|@]|9|@]|A|$]]|$1|D|3|E|5|6|7|N|8|@$F|O|G|P|H|I]|$F|Q|G|R|H|I]]|9|@]|A|$]]|$1|J|3|-4|5|6|7|S|8|@]|9|@]|A|$]]]|K|$]]

Do it only on the training set. And, of course, make sure that the augmentation does not make the label wrong (e.g. when rotating 6 and 9 by about 180°).

The reason why we use a training and a test set in the first place is that we want to estimate the error our system will have in reality. So the data for the test set should be as close to real data as possible.

If you do it on the test set, you might have the problem that you introduce errors. For example, say you want to recognize digits and you augment by rotating. Then a <code>6</code> might look like a <code>9</code>. But not all examples are that easy. Better be save than sorry.

blocks|key|608212|text|我认为，在某些情况下，为验证集使用数据增强可能会有帮助。|type|unstyled|depth|inlineStyleRanges|entityRanges|data|608213|例如，我训练了许多用于医学图像分割的CNN。许多我使用的增强转换的目的是降低图像质量，从而使网络能够对此类数据具有鲁棒性。如果训练集看起来很糟糕，并且验证集看起来很好，那么很难比较训练期间的损失，因此评估过强会很复杂。|offset|length|608214|除非我使用测试时增强来改进结果或估计任意不确定度，否则我永远不会对测试集使用增强。|608215|entityMap|0|LINK|mutability|MUTABLE|url|https://torchio.readthedocs.io/transforms/augmentation.html|1|https://www.sciencedirect.com/science/article/pii/S0925231219301961^0|0|O|8|0|0|G|8|1|0^^$0|@$1|2|3|4|5|6|7|R|8|@]|9|@]|A|$]]|$1|B|3|C|5|6|7|S|8|@]|9|@$D|T|E|U|1|V]]|A|$]]|$1|F|3|G|5|6|7|W|8|@]|9|@$D|X|E|Y|1|Z]]|A|$]]|$1|H|3|-4|5|6|7|10|8|@]|9|@]|A|$]]]|I|$J|$5|K|L|M|A|$N|O]]|P|$5|K|L|M|A|$N|Q]]]]

I would argue that, in some cases, using data augmentation for the validation set can be helpful.
For example, I train a lot of CNNs for medical image segmentation. Many of <a href="https://torchio.readthedocs.io/transforms/augmentation.html" rel="nofollow noreferrer">the augmentation transforms that I use</a> are meant to reduce the image quality so that the network is trained to be robust against such data. If the training set looks bad and the validation set looks nice, it will be hard to compare the losses during training and therefore assessing overfit will be complicated.
I would never use augmentation for the test set unless I'm using test-time augmentation to improve results or <a href="https://www.sciencedirect.com/science/article/pii/S0925231219301961" rel="nofollow noreferrer">estimate aleatoric uncertainty</a>.

blocks|key|1211107|text|在计算机视觉中，您可以在测试期间使用数据增强来获得关于测试图像的不同视图。然后，您必须聚合从每幅图像中获得的结果，例如，通过对它们进行平均处理。|type|unstyled|depth|inlineStyleRanges|offset|length|style|BOLD|entityRanges|data|1211108|例如，鉴于下面这个符号，改变观点可能导致不同的解释：|1211109|​|1211110|📷|atomic|1211111|1211112|entityMap|0|IMAGE|mutability|IMMUTABLE|imageUrl|https://developer.qcloudimg.com/http-save/yehe-900000/494d1ae7824d40d5c6ac2d20bdb067cb.png|imageAlt^0|N|8|0|0|0|0|1|0|0|0^^$0|@$1|2|3|4|5|6|7|W|8|@$9|X|A|Y|B|C]]|D|@]|E|$]]|$1|F|3|G|5|6|7|Z|8|@]|D|@]|E|$]]|$1|H|3|I|5|6|7|10|8|@]|D|@]|E|$]]|$1|J|3|K|5|L|7|11|8|@]|D|@$9|12|A|13|1|14]]|E|$]]|$1|M|3|I|5|6|7|15|8|@]|D|@]|E|$]]|$1|N|3|-4|5|6|7|16|8|@]|D|@]|E|$]]]|O|$P|$5|Q|R|S|E|$T|U|V|-4]]]]

In computer vision, you can use data augmentation during test time to obtain different views on the test image. You then have to aggregate the results obtained from each image for example by averaging them.

For example, given this symbol below, changing the point of view can lead to different interpretations : 

<a href="https://i.stack.imgur.com/2utvl.png" rel="nofollow noreferrer"><img src="https://i.stack.imgur.com/2utvl.png" alt="different view"></a>

blocks|key|608239|text|一些图像预处理软件工具，如Roboflow+(https://roboflow.com/)，也将数据增强应用于测试数据。我想说的是，如果一个人处理的是小而稀有的物体，比如大脑微出血(在磁共振图像上很小，很难辨认)，那么增强测试集可能是有用的。然后，您可以验证您的模型已经学会了在不同的方向和亮度条件下检测这些对象(假设您的培训数据已经以相同的方式得到了增强)。|type|unstyled|depth|inlineStyleRanges|entityRanges|offset|length|data|608240|entityMap|0|LINK|mutability|MUTABLE|url|https://roboflow.com/^0|N|L|0|0^^$0|@$1|2|3|4|5|6|7|L|8|@]|9|@$A|M|B|N|1|O]]|C|$]]|$1|D|3|-4|5|6|7|P|8|@]|9|@]|C|$]]]|E|$F|$5|G|H|I|C|$J|K]]]]

Some image preprocessing software tools like Roboflow (<a href="https://roboflow.com/" rel="nofollow noreferrer">https://roboflow.com/</a>) apply data augmentation to test data as well. I'd say that if one is dealing with small and rare objects, say, cerebral microbleeds (which are tiny and difficult to spot on magnetic resonance images), augmenting one's test set could be useful. Then you can verify that your model has learned to detect these objects given different orientation and brightness conditions (given that your training data has been augmented in the same way).

blocks|key|1211121|text|数据增强的目的是对模型进行推广，使其更好地了解图像的方向，从而使模型在测试过程中能够很好地理解测试数据。因此，仅在训练集中使用增强技术是很好的实践。|type|unstyled|depth|inlineStyleRanges|entityRanges|data|1211122|entityMap^0|0^^$0|@$1|2|3|4|5|6|7|D|8|@]|9|@]|A|$]]|$1|B|3|-4|5|6|7|E|8|@]|9|@]|A|$]]]|C|$]]

The goal of data augmentation is to generalize the model and make it learn more orientation of the images, such that the during testing the model is able to apprehend the test data well. So, it is well practiced to use augmentation technique only for training sets.

blocks|key|607478|text|添加验证数据的目的是建立广义模型，因此它只不过是预测真实世界的数据。为了预测真实世界的数据，验证集应该包含真实数据.增加验证数据没有问题，但不会提高模型的准确性。|type|unstyled|depth|inlineStyleRanges|entityRanges|data|607479|entityMap^0|0^^$0|@$1|2|3|4|5|6|7|D|8|@]|9|@]|A|$]]|$1|B|3|-4|5|6|7|E|8|@]|9|@]|A|$]]]|C|$]]

The point of adding validation data is to build generalized model so it is nothing but to predict real-world data. inorder to predict real-world data, the validation set should contain real data. There is no problem with augmenting validation data but it won't increase the accuracy of the model.

blocks|key|607501|text|这是我的两分钱：|type|unstyled|depth|inlineStyleRanges|entityRanges|data|607502|您可以在培训数据和(验证数据)上训练您的模型:前者优化您的参数，后者给您一个适当的停止条件。测试数据是给您一个真实的估计，您可以期望您的模型的表现。|offset|length|style|BOLD|607503|对于培训，您可以增强您的培训数据，以增强对各种因素的鲁棒性，包括但不限于采样误差、数据源之间的偏差、全局数据分布、定位以及任何其他您想要考虑的变化。|607504|当模型最具通用性时，验证数据应指示给训练方法。按照这种逻辑，如果您希望看到一些可以通过数据增强来模拟的真实世界数据的变化，那么无论如何，验证数据集都应该是增强的。|607505|另一方面，测试数据不应该扩展，除非在数据非常有限的特殊情况下，而且对测试数据的真实性能的估计有太多的差异。|607506|entityMap^0|0|8|1|0|8|8|0|1W|C|0|5|4|0^^$0|@$1|2|3|4|5|6|7|P|8|@]|9|@]|A|$]]|$1|B|3|C|5|6|7|Q|8|@$D|R|E|S|F|G]]|9|@]|A|$]]|$1|H|3|I|5|6|7|T|8|@$D|U|E|V|F|G]]|9|@]|A|$]]|$1|J|3|K|5|6|7|W|8|@$D|X|E|Y|F|G]]|9|@]|A|$]]|$1|L|3|M|5|6|7|Z|8|@$D|10|E|11|F|G]]|9|@]|A|$]]|$1|N|3|-4|5|6|7|12|8|@]|9|@]|A|$]]]|O|$]]

Here are my two cents:
You train your model on the training data and the validation data: the former to optimize your parameters, and the latter to give you an appropriate stopping condition. The test data is to give you a real-world estimate of how well you can expect your model to perform.
For training, you can augment your training data to increase robustness to various factors including, but not limited to, sampling error, bias between data sources, shifts in global data distribution, positioning, and any other sort of variation you would like to account for.
The validation data should indicate to the training method when the model is most generalizable. By this logic, if you expect to see some variation in real-world data that can be simulated using data augmentation, then by all means, the validation dataset should be augmented.
The test data, on the other hand, should not be augmented, except potentially in special scenarios where data is very limited, and an estimate of real-world performance on test data has too much variance.

blocks|key|607521|text|可以在培训、验证和测试集中使用增强数据。|type|unstyled|depth|inlineStyleRanges|entityRanges|data|607522|唯一要避免的是在验证或测试集中使用来自培训集的相同数据。|607523|例如，如果您从培训数据的寄存器中生成3个增广实例，请确保这3个增广实例中没有一个意外地在验证或测试集中结束。|607524|事实证明，使用来自训练集的数据，甚至是增强的数据，来验证或测试一个模型是一个方法论错误。|607525|entityMap^0|0|0|0|0^^$0|@$1|2|3|4|5|6|7|J|8|@]|9|@]|A|$]]|$1|B|3|C|5|6|7|K|8|@]|9|@]|A|$]]|$1|D|3|E|5|6|7|L|8|@]|9|@]|A|$]]|$1|F|3|G|5|6|7|M|8|@]|9|@]|A|$]]|$1|H|3|-4|5|6|7|N|8|@]|9|@]|A|$]]]|I|$]]

You can use augmentation data in training, validation and test sets.
The only thing to avoid is using the same data from the training set in validation or test sets.
For example, if you generate 3 augmented instances from an register of the training data, make sure that no one of these 3 augmented instances accidentally ends up in the validation or test sets.
It turns out that using data from the training set, even augmented data, to validate or test a model is a methodology mistake.

It is common practice to augment data (add samples programmatically, such as random crops, etc. in the case of a dataset consisting of images) on both training and test set, or just the training data set?

Data augmentation in test/validation set?

翻译质量差，导致语言生硬或混乱。

没有提供实际的解决方法或示例。

解答不清晰，无法理解或解决问题。

页面排版不美观，阅读体验差。

文章

问答

视频

学习中心

腾讯云实验室

直播

竞赛

腾讯云代码分析专区

腾讯iOA零信任安全管理系统专区

腾讯云架构师技术同盟交流圈

腾讯云数据库专区

腾讯云智能顾问专区

腾讯云原生专区

腾讯混元专区

腾讯云TCE专区

腾讯云Lighthouse专区

腾讯云HAI专区

腾讯云Edgeone专区

腾讯云存储专区

腾讯云智能专区

腾讯轻联专区 

腾讯云开发专区

TAPD专区

腾讯轻量云游戏服专区

腾讯云最具价值专家

腾讯云架构师技术同盟

腾讯云创作之星

腾讯云开发者先锋

腾讯云代码助手

云原生构建

TAPD 敏捷项目管理

Cloud Studio

SDK中心

API中心

命令行工具

涵盖代码开发、场景应用、自动测试全流程，助你从零构建专属AI助手

一站式MCP教程库，解锁AI应用新玩法

聚焦“写作效率、视觉美观与运行性能”三方面进行全面升级，为您提供更高效、稳定的创作环境

社区富文本&Markdown编辑器全新改版上线，欢迎大家体验!

诚挚邀请您参与本次调研，分享您的真实使用感受与建议。您的反馈至关重要，感谢您的支持与参与！

社区新版编辑器体验调研

通常的做法是在训练和测试集，或者仅仅是训练数据集上增加数据(在包含图像的数据集中，以编程方式添加样本，例如随机作物等)？

问测试/验证集中的数据增强？
EN

回答 11

Stack Overflow用户

Stack Overflow用户

Stack Overflow用户

社区

活动

圈层

关于

腾讯云开发者

热门产品

热门推荐

更多推荐

问测试/验证集中的数据增强？EN

回答 11

Stack Overflow用户

Stack Overflow用户

Stack Overflow用户

社区

活动

圈层

关于

腾讯云开发者

热门产品

热门推荐

更多推荐

问测试/验证集中的数据增强？
EN