Towards Efficient Cross-Modal Anomaly Detection Using Triple-Adaptive Network and Bi-Quintuple Contrastive Learning | IEEE Journals & Magazine | IEEE Xplore