提出可解释图像分类的因果解释框架，兼具形式严谨与黑盒可计算性。
Sufficient, Necessary and Complete Causal Explanations in Image Classification
- 用因果逻辑定义解释，区分充分、必要与完整成分
- 在ResNet上平均6秒完成全类型解释，无需模型内部信息
- 适用于不同模型，揭示其解释模式差异，适合可信AI研究者
现有图像分类解释方法缺乏形式化基础，而基于逻辑的解释虽严谨但依赖强假设。本文证明因果解释在形式上等价于逻辑解释，同时支持黑盒算法且自然适配图像分类模型。我们提出δ-完备解释（最低置信阈值）和1-完备因果解释（与原图同置信）。通过实验发现不同模型具有不同的充分性、必要性与完备性模式。算法高效可计算，平均6秒/图像，完全黑盒，无需模型内部知识、梯度或单调性等性质。
原文摘要 · Abstract (English)
Existing algorithms for explaining the outputs of image classifiers are based on a variety of approaches and produce explanations that frequently lack formal rigour. On the other hand, logic-based explanations are formally and rigorously defined but their computability relies on strict assumptions about the model that do not hold on image classifiers. In this paper, we show that causal explanations, in addition to being formally and rigorously defined, enjoy the same formal properties as logic-based ones, while still lending themselves to black-box algorithms and being a natural fit for image classifiers. We prove formal properties of causal explanations and their equivalence to logic-based explanations. We demonstrate how to subdivide an image into its sufficient and necessary components. We introduce $δ$-complete explanations, which have a minimum confidence threshold and 1-complete causal explanations, explanations that are classified with the same confidence as the original image. We implement our definitions, and our experimental results demonstrate that different models have different patterns of sufficiency, necessity, and completeness. Our algorithms are efficiently computable, taking on average 6s per image on a ResNet model to compute all types of explanations, and are totally black-box, needing no knowledge of the model, no access to model internals, no access to gradient, nor requiring any properties, such as monotonicity, of the model.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。