RESEARCHarXiv CS.LG·14d ago
CAFD: Concept-Aware DNN Fault Detection using VLMs
CAFD is a new learning-based method for detecting faults in Deep Neural Networks (DNNs) that combines multiple information sources for superior performance and efficiency. It utilizes model-based signals, distance features, and a novel Concept Failure Ratio (CFR) derived from Vision-Language Models (VLMs).
27