Visual Concept Networks: A Graph-Based Approach to Detecting Anomalous Data in Deep Neural Networks
This addresses robustness issues in DNNs for applications dealing with complex real-world anomalies, though it appears incremental as it builds on existing OOD detection methods with a new approach.
The paper tackles the problem of deep neural networks struggling with anomalous and out-of-distribution data by introducing a graph-based method using visual concept networks, demonstrating effectiveness in detecting both far-OOD and near-OOD data through extensive testing on novel tasks.
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.