https://phuchoang2603.github.io/sis-arxiv-vad-papers/authors/vu-hoang-minh-doan/Recent content in Vu Hoang Minh Doan on sis-arxiv-vad-papersHugo -- gohugo.ioenMon, 31 Mar 2025 00:00:00 +0000https://phuchoang2603.github.io/sis-arxiv-vad-papers/papers/aadc-net_a_multimodal_deep_learning_framework_for_automatic_anomaly_detection_in_real-time_surveillance/Mon, 31 Mar 2025 00:00:00 +0000https://phuchoang2603.github.io/sis-arxiv-vad-papers/papers/aadc-net_a_multimodal_deep_learning_framework_for_automatic_anomaly_detection_in_real-time_surveillance/Introduces AADC-Net, a multimodal deep neural network leveraging pretrained vision-language models, large language models, and object detection (DETR) for real-time anomaly detection and categorization in surveillance videos. The framework addresses data scarcity, imbalance, and computational challenges, demonstrating state-of-the-art performance on multiple datasets, with practical deployment in smart gyms and healthcare settings.