mlsanityData to 7 Oct 2026

Cerberus: Real-Time Video Anomaly Detection via Cascaded Vision-Language Models

arXiv 2510.16290 ↗First posted 18 Oct 2025cs.CV cs.CLcited by 1

On the map: Multimodal LLMs / vision-language models → Chart understanding and anomaly detection

Abstract (arXiv)

Video anomaly detection (VAD) has rapidly advanced by recent development of Vision-Language Models (VLMs). While these models offer superior zero-shot detection capabilities, their immense computational cost and unstable visual grounding performance hinder real-time deployment. To overcome these challenges, we introduce Cerberus, a two-stage cascaded system designed for efficient yet accurate real-time VAD. Cerberus learns normal behavioral rules offline, and combines lightweight filtering with fine-grained VLM reasoning during online inference. The performance gains of Cerberus come from two key innovations: motion mask prompting and rule-based deviation detection. The former directs the VLM's attention to regions relevant to motion, while the latter identifies anomalies as deviations from learned norms rather than enumerating possible anomalies. Extensive evaluations on four datasets show that Cerberus on average achieves 57.68 fps on an NVIDIA L40S GPU, a 151.79× speedup, and 97.2% accuracy comparable to the state-of-the-art VLM-based VAD methods, establishing it as a practical solution for real-time video analytics.

Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Placed in Chart understanding and anomaly detection, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 27 of its 51 references that are in the map. First posted on arXiv 18 Oct 2025.

Builds on

27 of 51 references.

Closest ancestors · 1

Same direction, newest first

  1. 2024Do LLMs Understand Visual Anomalies? Uncovering LLM's Capabilities in Zero-shot Anomaly Detection15 Apr

Other references · 26

  1. 2017Attention Is All You Needcited by 59,932
  2. 2023GPT-4 Technical Reportcited by 18,132
  3. 2017Quo Vadis, Action Recognition? A New Model and the Kinetics Datasetcited by 3,259
  4. 2023Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyondcited by 2,179
  5. 2025Qwen3 Embedding: Advancing Text Embedding and Reranking Through Foundation Modelscited by 1,175
  6. 2018Real-world Anomaly Detection in Surveillance Videoscited by 570
  7. 2024YOLOv10: Real-Time End-to-End Object Detectioncited by 508
  8. 2025Perception Encoder: The best visual embeddings are not at the output of the networkcited by 315
  9. 2020Not only Look, but also Listen: Learning Multimodal Violence Detection under Weak Supervisioncited by 235
  10. 2021Weakly-supervised Video Anomaly Detection with Robust Temporal Feature Magnitude Learningcited by 226
  11. 2022mPLUG: Effective and Efficient Vision-Language Learning by Cross-modal Skip-connectionscited by 185
  12. 2021A Hybrid Video Anomaly Detection Framework via Memory-Augmented Flow Reconstruction and Flow-Guided Frame Predictioncited by 129
14 more
  1. 2024A Review on Edge Large Language Models: Design, Execution, and Applicationscited by 98
  2. 2020A Survey of Single-Scene Video Anomaly Detectioncited by 88
  3. 2023Fast and Robust Early-Exiting Framework for Autoregressive Language Models with Synchronized Parallel Decodingcited by 80
  4. 2024VideoVista: A Versatile Benchmark for Video Understanding and Reasoningcited by 72
  5. 2019GODS: Generalized One-class Discriminative Subspaces for Anomaly Detectioncited by 64
  6. 2020Ekya: Continuous Learning of Video Analytics Models on Edge Compute Serverscited by 58
  7. 2024Hawk: Learning to Understand Open-World Video Anomaliescited by 47
  8. 2020Robust Unsupervised Video Anomaly Detection by Multi-Path Frame Predictioncited by 42
  9. 2023A New Comprehensive Benchmark for Semi-supervised Video Anomaly Detection and Anticipationcited by 42
  10. 2024Weakly Supervised Video Anomaly Detection and Localization with Spatio-Temporal Promptscited by 40
  11. 2021A Vision-based System for Traffic Anomaly Detection using Deep Learning and Decision Treescited by 26
  12. 2024An Exploratory Study on Human-Centric Video Anomaly Detection through Variational Autoencoders and Trajectory Predictioncited by 19
  13. 2025Sherlock: Towards Multi-scene Video Abnormal Event Extraction and Localization via a Global-local Spatial-sensitive LLMcited by 10
  14. 2025AVA: Towards Agentic Video Analytics with Vision Language Modelscited by 5

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026NOVA: Normal-Side Modeling for Training-Free Zero-Shot Video Anomaly Detectioncited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025SlowFastVAD: Video Anomaly Detection via Integrating Simple Detector and RAG-Enhanced Vision-Language Model
  2. 2024Follow the Rules: Reasoning for Video Anomaly Detection with Large Language Models
  3. 2026GridVAD: Open-Set Video Anomaly Detection via Spatial Reasoning over Stratified Frame Grids
  4. 2024Quo Vadis, Anomaly Detection? LLMs and VLMs in the Spotlight
  5. 2026PrismVAU: Prompt-Refined Inference System for Multimodal Video Anomaly Understanding
  6. 2024Video Anomaly Detection in 10 Years: A Survey and Outlook
  7. 2026Is Video Anomaly Detection Misframed? Evidence from LLM-Based and Multi-Scene Models
  8. 2026CoReVAD: A Contextual Reasoning Framework for Training-Free Video Anomaly Detection