不同质量数据下分类器学习的特征范数压缩映射
Contraction Mapping of Feature Norms for Classifier Learning on the Data with Different Quality
摘要：
流行的softmax loss及其最近的扩展在基于深度学习的图像分类中取得了巨大的成功。然而，用于训练图像分类器的数据通常具有不同的质量。如果不考虑这一问题，低质量数据的正确分类就难以解决。在本文中，我们通过对各种应用和各种深度神经网络的仔细实验，发现了图像的特征范数与其质量之间的正相关性。基于这一发现，我们提出了一个收缩映射函数，根据训练图像的质量来压缩训练图像的特征范数范围，并将此收缩映射函数嵌入到softmax loss或其扩展中，以产生新的学习目标。在手写体数字识别、肺结节分类、人脸验证和人脸识别等分类应用中的实验表明，该方法能够有效地处理不同质量数据的学习问题，并在分类精度上有显著而稳定的提高。
英文摘要：
The popular softmax loss and its recent extensions have achieved great success in the deep learning-based image classification. However, the data for training image classifiers usually has different quality. Ignoring such problem, the correct classification of low quality data is hard to be solved. In this paper, we discover the positive correlation between the feature norm of an image and its quality through careful experiments on various applications and various deep neural networks. Based on this finding, we propose a contraction mapping function to compress the range of feature norms of training images according to their quality and embed this contraction mapping function into softmax loss or its extensions to produce novel learning objectives. The experiments on various classification applications, including handwritten digit recognition, lung nodule classification, face verification and face recognition, demonstrate that the proposed approach is promising to effectively deal with the problem of learning on the data with different quality and leads to the significant and stable improvements in the classification accuracy.
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。
