AIbase
EN
Home
Violence Detect 44
V

Violence Detect 44

Developed by Hemg
An audio classification model fine-tuned from facebook/wav2vec2-base-960h for detecting violent sounds
Audio Classification
Transformers
Open Source License:Apache-2.0#Audio Classification#Violence Detection#High Accuracy
Downloads 28
Release Time : 4/3/2024

Model Overview

This model is an audio classification model based on the wav2vec2 architecture, specifically designed to recognize and classify violent sounds. It achieved an accuracy of 89.33% on the evaluation dataset.

Model Features

High Accuracy
Achieves 89.33% classification accuracy on the evaluation dataset
Based on wav2vec2 Architecture
Fine-tuned using the proven wav2vec2-base-960h architecture
Lightweight
Based on the base model, suitable for deployment in resource-constrained environments

Model Capabilities

Audio Classification
Violence Sound Detection
Audio Feature Extraction

Use Cases

Security Monitoring
Public Space Violence Detection
Detects potential violent incidents in public spaces through audio analysis
Can identify 89.33% of violent sounds in real-time
Smart Home
Home Security Alerts
Detects abnormal violent sounds in home environments and triggers alarms
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
EnglishįŽ€äŊ“中文įšéĢ”中文ãĢãģんご
Š 2025AIbase