Videomae Base Finetuned Ucf Crime
A video analysis model fine-tuned on the UCF-Crime dataset based on the VideoMAE foundation model
Downloads 18
Release Time : 5/21/2023
Model Overview
This model is a video understanding model based on the VideoMAE architecture, specifically fine-tuned for crime detection scenarios, capable of analyzing video content and identifying potential criminal behaviors or abnormal activities.
Model Features
Video Anomaly Detection
Video analysis capability optimized specifically for crime scenarios
Based on Self-Supervised Learning
Utilizes VideoMAE's masked autoencoder pre-training method
Lightweight Fine-tuning
Efficient fine-tuning on the foundation model to adapt to specific tasks
Model Capabilities
Video Content Analysis
Anomaly Behavior Detection
Crime Scene Recognition
Use Cases
Public Safety
Surveillance Video Analysis
Automatically analyzes suspicious behaviors in surveillance videos
Accuracy 37.2%
Anomalous Event Detection
Identifies violent, theft, and other anomalous events in videos
Featured Recommended AI Models
Š 2025AIbase