Context-Aware Video Anomaly Detection in Long-Term Datasets with Temporal and Spatial Contextual Awareness
The core message of this work is to propose a context-aware video anomaly detection algorithm, Trinity, that can effectively detect anomalies in long-term video datasets by learning alignments between video content (appearance and motion) and contextual information (time of day, day of week, game schedule, etc.).
Video Anomaly Detection via Spatio-Temporal Pseudo-Anomaly Generation: A Unified Approach
A novel method for generating generic spatio-temporal pseudo-anomalies by inpainting masked regions using a pre-trained latent diffusion model and perturbing optical flow using mixup. This unified framework measures reconstruction quality, temporal irregularity, and semantic inconsistency to effectively detect real-world anomalies.
Dynamic Distinction Learning: Adaptive Pseudo Anomalies for Enhancing Video Anomaly Detection
A novel video anomaly detection methodology that combines pseudo-anomalies, dynamic anomaly weighting, and a distinction loss function to improve detection accuracy by adaptively learning the optimal balance between normal and anomalous patterns.
Leveraging Large Language Models and Vision-Language Models for Training-free Video Anomaly Detection
A novel training-free method for video anomaly detection that leverages pre-trained large language models and vision-language models to detect anomalies without any task-specific training or data collection.
Collaborative Learning of Anomalies with Privacy (CLAP): A New Baseline for Unsupervised Video Anomaly Detection
CLAP is a new baseline for unsupervised video anomaly detection that enables collaborative training of anomaly detection models across multiple participants without compromising data privacy.
MULDE: Multiscale Log-Density Estimation for Video Anomaly Detection
ビデオの異常検出における新しいアプローチを提案する。
GlanceVAD: Label-efficient Video Anomaly Detection with Glance Supervision
Glance annotation paradigm improves anomaly detection efficiency and model performance in video anomaly detection.
Open-Vocabulary Video Anomaly Detection: Leveraging Large Models for Improved Detection and Categorization
Exploring OVVAD with large pre-trained models enhances anomaly detection and categorization.
GlanceVAD: Label-efficient Video Anomaly Detection with Glance Supervision
The author introduces a novel labeling paradigm, "glance annotation," for video anomaly detection to balance accuracy and cost-effectiveness. By leveraging Gaussian kernels through Temporal Gaussian Splatting, the GlanceVAD method outperforms existing approaches in model performance and annotation efficiency.
Learning Suspected Anomalies from Event Prompts for Video Anomaly Detection
The author proposes a novel framework to guide the learning of suspected anomalies from event prompts, enhancing weakly supervised video anomaly detection. By utilizing semantic anomaly similarity and multi-prompt learning, the model outperforms state-of-the-art methods in various datasets.
© 2024 by Linnk AI