Research map / Video / Video segmentation and tracking
Generalizing Single-Frame Supervision to Event-Level Understanding for Video Anomaly Detection
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes Single-Frame supervised Video Anomaly Detection (SF-VAD), a new paradigm that uses one annotated frame per anomalous event instead of full or weak video-level labels.
Problem
Video anomaly detection suffers from heavy annotation cost in fully supervised settings, insensitivity to subtle anomalies in semi-supervised settings, and noise vulnerability in weakly supervised settings.
Approach
Replaces full or video-level supervision with a single annotated frame per event and generalizes it to event-level understanding for detecting abnormal frames from discrete events in video.
Closest papers by meaning
- Learning Event Completeness for Weakly Supervised Video Anomaly Detection (ICML 2025 · similarity 0.91)
- Open-Vocabulary Video Anomaly Detection (CVPR 2024 · similarity 0.90)
- Harnessing Large Language Models for Training-free Video Anomaly Detection (CVPR 2024 · similarity 0.90)
- VADTree: Explainable Training-Free Video Anomaly Detection via Hierarchical Granularity-Aware Tree (NeurIPS 2025 · similarity 0.90)
- MoniTor: Exploiting Large Language Models with Instruction for Online Video Anomaly Detection (NeurIPS 2025 · similarity 0.89)
- Holmes-VAU: Towards Long-term Video Anomaly Understanding at Any Granularity (CVPR 2025 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.