---
title: "MemoVAD: Resource-Efficient Video Anomaly Detection via Dynamic Semantic Memory in Edge Computing Scenarios"
authors:
  - "Guo Li"
  - "Jiandian Zeng"
  - "Yang Li"
  - "Zihao Peng"
  - "Ke Chen"
  - "Tian Wang"
doi: "10.24963/ijcai.2026/618"
url: "https://ijcai.org/proceedings/2026/618"
pdf: "https://ijcai.org/proceedings/2026/0618.pdf"
venue: "Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence"
year: 2026
pages: "5549-5557"
track: "Main Track"
keywords:
  - "Multidisciplinary Topics and Applications: Ubiquitous computing cystems"
  - "Multidisciplinary Topics and Applications: Real-time systems"
  - "Multidisciplinary Topics and Applications: Sensor networks and smart cities"
---

# MemoVAD: Resource-Efficient Video Anomaly Detection via Dynamic Semantic Memory in Edge Computing Scenarios

Guo Li, Jiandian Zeng, Yang Li, Zihao Peng, Ke Chen, Tian Wang

*Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence*, pages 5549-5557. DOI: [10.24963/ijcai.2026/618](https://doi.org/10.24963/ijcai.2026/618)

## Abstract

Deploying Video Anomaly Detection (VAD) in real-world surveillance faces a fundamental tension between the demand for high-level semantics to ensure effectiveness and the limited computational resources of edge devices. Vision–Language Models (VLMs) provide rich open-vocabulary semantics, but their latency and computational cost preclude on-device deployment. To address the challenge, we propose MemoVAD, an edge–cloud collaborative framework that selectively incorporates VLM semantics into streaming VAD. MemoVAD runs most inference on the edge with a lightweight detector and a causal Temporal Context Encoder (TCE) to model temporal dependencies. Specifically, we introduce an Uncertainty-Aware Gating (UAG) policy grounded in Subjective Logic to model perceived uncertainty and query the cloud-based VLM only for high-uncertainty and semantically novel clips. Besides, a Dynamic Semantic Memory (DSM) is designed to cache VLM-verified prototypes for efficient retrieval, enabling the edge model to progressively incorporate VLM-level semantics via a semantic adapter. Experiments on UCF-Crime and XD-Violence datasets via a real edge device show that MemoVAD substantially reduces communication overhead while surpassing state-of-the-art performance. The demo video is available at: https://memovad2026.github.io/.