用图像的视觉概念图检测异常数据，提升模型鲁棒性。
Visual Concept Networks: A Graph-Based Approach to Detecting Anomalous Data in Deep Neural Networks
- 将图像转为可理解的视觉概念图，利用图结构分析异常
- 在新任务上验证，对远距离和近距离异常数据均有效
- 适合关注模型安全与真实场景泛化能力的研究者
深度神经网络（DNN）在诸多应用中日益普及，但在面对异常和分布外（OOD）数据时仍缺乏鲁棒性。现有OOD评测常过于简化，聚焦单物体任务，未能充分反映真实世界中的复杂异常。本文提出一种新颖且直接的方法，通过图结构与拓扑特征有效检测远距离与近距离的OOD数据。将图像转换为由可解释的视觉概念构成的互联网络，通过在两个新任务上的大量测试，包括使用大规模词汇表和多样化任务的消融研究，证明了该方法的有效性。该方法增强了DNN对OOD数据的抗性，有望在多种应用场景中提升性能。
原文摘要 · Abstract (English)
Deep neural networks (DNNs), while increasingly deployed in many applications, struggle with robustness against anomalous and out-of-distribution (OOD) data. Current OOD benchmarks often oversimplify, focusing on single-object tasks and not fully representing complex real-world anomalies. This paper introduces a new, straightforward method employing graph structures and topological features to effectively detect both far-OOD and near-OOD data. We convert images into networks of interconnected human understandable features or visual concepts. Through extensive testing on two novel tasks, including ablation studies with large vocabularies and diverse tasks, we demonstrate the method's effectiveness. This approach enhances DNN resilience to OOD data and promises improved performance in various applications.
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。