arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2309.14531cs.CV
收藏

像素锚定的原型部件网络

Pixel-Grounded Prototypical Part Networks

University of Notre Dame(圣母大学);Mitsubishi Electric Research Laboratories(三菱电机研究实验室)

Zachariah Carmichael, Suhas Lohit, Anoop Cherian, Michael Jones, Walter Scheirer

AI总结:

针对 PROTOPNET 类原型部件网络实际定位整图而非部件的问题,本文提出 PIXPNET,通过感受野约束、像素空间映射和架构修正实现真实部件定位,并在保持准确率的同时提升可解释性。

摘要:

原型部件神经网络(ProtoPartNN),即 PROTOPNET 及其衍生模型,是一种本质上可解释的机器学习方法。它们的原型学习机制能够产生直观解释,其形式为:这个(原型)看起来像那个(测试图像块)。但是,这真的看起来像那吗?在本研究中,我们深入考察了以往工作中对象部件定位及相关热力图为何具有误导性。现有 ProtoPartNN 并未定位到对象部件,而是定位到整幅图像,这与其生成的解释性可视化结果相反。我们认为,这些底层问题之所以被忽视,是因为可视化具有迷惑性,并且人们过度依赖直觉。为缓解这些问题,我们设计了基于感受野的新型架构约束以实现有意义的定位,并为 ProtoPartNN 提出了有原则的像素空间映射。为提升可解释性,我们还提出了额外的架构改进,包括简化的分类头。我们也对 PROTOPNET 及其衍生模型作出了其他修正,例如在训练期间使用验证集而非测试集来评估泛化能力。我们的方法 PIXPNET(Pixel-grounded Prototypical part Network,像素锚定原型部件网络)是唯一能够真正学习并定位原型对象部件的 ProtoPartNN。我们证明,PIXPNET 在不牺牲准确率的情况下,实现了可量化提升的可解释性。

英文摘要:

Prototypical part neural networks (ProtoPartNNs), namely PROTOPNET and its derivatives, are an intrinsically interpretable approach to machine learning. Their prototype learning scheme enables intuitive explanations of the form, this (prototype) looks like that (testing image patch). But, does this actually look like that? In this work, we delve into why object part localization and associated heat maps in past work are misleading. Rather than localizing to object parts, existing ProtoPartNNs localize to the entire image, contrary to generated explanatory visualizations. We argue that detraction from these underlying issues is due to the alluring nature of visualizations and an over-reliance on intuition. To alleviate these issues, we devise new receptive field-based architectural constraints for meaningful localization and a principled pixel space mapping for ProtoPartNNs. To improve interpretability, we propose additional architectural improvements, including a simplified classification head. We also make additional corrections to PROTOPNET and its derivatives, such as the use of a validation set, rather than a test set, to evaluate generalization during training. Our approach, PIXPNET (Pixel-grounded Prototypical part Network), is the only ProtoPartNN that truly learns and localizes to prototypical object parts. We demonstrate that PIXPNET achieves quantifiably improved interpretability without sacrificing 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。

↑