arXiv 2301.040112023-10-22cs.CV

学习支持原型和平凡原型以实现可解释的图像分类

Learning Support and Trivial Prototypes for Interpretable Image Classification

Australian Institute for Machine Learning, University of Adelaide（阿德莱德大学澳大利亚机器学习研究所）；Harvard University（哈佛大学）；St Vincent’s Institute of Medical Research（圣文森特医学研究所）；St Vincent’s Hospital Melbourne（墨尔本圣文森特医院）；CVSSP, University of Surrey（萨里大学计算机视觉、语音与信号处理中心）

Chong Wang, Yuyuan Liu, Yuanhong Chen, Fengbei Liu, Yu Tian, Davis J. McCarthy, Helen Frazer, Gustavo Carneiro

备注：: ICCV 2023, Code: https://github.com/cwangrun/ST-ProtoPNet

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

提出学习靠近分类边界的支持原型，结合平凡原型构建 ST-ProtoPNet 模型，提升了图像分类的准确率与可解释性。

摘要：

原型部分网络（ProtoPNet）方法旨在通过将预测与一组训练原型相关联来实现可解释的分类，我们将这些训练原型称为平凡原型，因为它们在特征空间中被训练为远离分类边界。注意，可以在 ProtoPNet 和支持向量机（SVM）之间进行类比，因为这两种方法的分类都依赖于计算与一组训练点（即 ProtoPNet 中的平凡原型和 SVM 中的支持向量）的相似度。然而，平凡原型位于远离分类边界的地方，而支持向量则位于靠近该边界的地方，我们认为这种与成熟的 SVM 理论的差异可能导致 ProtoPNet 模型的分类准确率较低。在本文中，我们旨在通过一种学习支持原型的新方法来改进 ProtoPNet 的分类，正如 SVM 理论所建议的，这些支持原型位于特征空间中靠近分类边界的地方。此外，我们通过一个名为 ST-ProtoPNet 的新模型来进一步改善分类结果，该模型利用我们提出的支持原型和平凡原型来提供更有效的分类。在 CUB-200-2011、Stanford Cars 和 Stanford Dogs 数据集上的实验结果表明，ST-ProtoPNet 实现了最先进的分类准确率和可解释性结果。我们还表明，所提出的支持原型往往能更好地定位在目标物体上，而不是背景区域。

英文摘要：

Prototypical part network (ProtoPNet) methods have been designed to achieve interpretable classification by associating predictions with a set of training prototypes, which we refer to as trivial prototypes because they are trained to lie far from the classification boundary in the feature space. Note that it is possible to make an analogy between ProtoPNet and support vector machine (SVM) given that the classification from both methods relies on computing similarity with a set of training points (i.e., trivial prototypes in ProtoPNet, and support vectors in SVM). However, while trivial prototypes are located far from the classification boundary, support vectors are located close to this boundary, and we argue that this discrepancy with the well-established SVM theory can result in ProtoPNet models with inferior classification accuracy. In this paper, we aim to improve the classification of ProtoPNet with a new method to learn support prototypes that lie near the classification boundary in the feature space, as suggested by the SVM theory. In addition, we target the improvement of classification results with a new model, named ST-ProtoPNet, which exploits our support prototypes and the trivial prototypes to provide more effective classification. Experimental results on CUB-200-2011, Stanford Cars, and Stanford Dogs datasets demonstrate that ST-ProtoPNet achieves state-of-the-art classification accuracy and interpretability results. We also show that the proposed support prototypes tend to be better localised in the object of interest rather than in the background region.

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。