arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1909.05904eess.IVcs.CV
收藏

感知图像异常检测

Perceptual Image Anomaly Detection

Philips Research(飞利浦研究院);Lomonosov Moscow State University(莫斯科国立罗蒙诺索夫大学)

Nina Tuluptceva, Bart Bakker, Irina Fedulova, Anton Konushin

摘要:

我们提出了一种新的图像异常检测方法,其中使用的算法,从一些分布的“正常”数据的样本,旨在检测出的分布(异常)样本。我们的方法包括编码器和生成器的组合,用于将图像分布映射到预定义的潜在分布,反之亦然。它利用生成对抗网络来学习这些数据分布,并使用感知损失来检测图像异常。为了实现这一目标,我们引入了一个新的相似性度量,它表示图像之间的感知相似性,并且对图像对比度的变化具有鲁棒性。其次,我们介绍了一种新的方法来选择的权重的多目标损失函数(图像重建和分布映射)的情况下,验证数据集的超参数调整。在训练之后,我们的模型将输入图像的异常性测量为它与建模数据分布的最接近的生成图像之间的感知差异。所提出的方法进行了广泛的评估,在几个公开的图像基准,并实现了最先进的性能。

英文摘要:

We present a novel method for image anomaly detection, where algorithms that use samples drawn from some distribution of "normal" data, aim to detect out-of-distribution (abnormal) samples. Our approach includes a combination of encoder and generator for mapping an image distribution to a predefined latent distribution and vice versa. It leverages Generative Adversarial Networks to learn these data distributions and uses perceptual loss for the detection of image abnormality. To accomplish this goal, we introduce a new similarity metric, which expresses the perceived similarity between images and is robust to changes in image contrast. Secondly, we introduce a novel approach for the selection of weights of a multi-objective loss function (image reconstruction and distribution mapping) in the absence of a validation dataset for hyperparameter tuning. After training, our model measures the abnormality of the input image as the perceptual dissimilarity between it and the closest generated image of the modeled data distribution. The proposed approach is extensively evaluated on several publicly available image benchmarks and achieves state-of-the-art performance.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑