Scribd
Upload
10 views9 pages

Suspicious-Human - Activity - Recognition Final

The document presents a project on Real-Time Violence Detection using ViT and LSTM, aimed at automating the detection of suspicious activities in crowded areas through computer vision. The system is designed to enhance surveillance efficiency, reduce human dependency, and can be integrated with existing infrastructure. Key features include live monitoring, the use of deep learning algorithms, and a workflow that involves preprocessing and activity recognition.

Uploaded by

aishwaryabajntri3
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
10 views9 pages

Suspicious-Human - Activity - Recognition Final

The document presents a project on Real-Time Violence Detection using ViT and LSTM, aimed at automating the detection of suspicious activities in crowded areas through computer vision. The system is designed to enhance surveillance efficiency, reduce human dependency, and can be integrated with existing infrastructure. Key features include live monitoring, the use of deep learning algorithms, and a workflow that involves preprocessing and activity recognition.

Uploaded by

aishwaryabajntri3
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 9

ANGADI INSTITUTE OF TECHNOLOGY AND

MANAGEMENT, BELAGAVI - 590009


Department of Computer Science
and Engineering

Real-Time Violence Detection using ViT + LSTM

Presented By: Guided By:


Mr. VIDYADHAR HANJI
Mr. Amit Rangarava Ingale Assistant Professor
Mr. Karan Chandrakant Patil Dept of CSE, AITM
INDEX
• Introduction
• Objectives
• Key Features
• Workflow
• System Architecture Diagram
• Conclusion

Dept. of CSE,AITM 2
INTRODUCTION
• Problem: Monitoring Crowded Area to prevent unusual activities is
challenging due to human limitations.
• Solution: The project uses computer vision to automate the
detection of suspicious activities in video feeds, enhancing
surveillance efficiency.
• Applications: Ideal for public spaces like malls, schools, and
parking lots…etc.

Dept. of CSE,AITM 3
OBJECTIVES
 To develop a monitoring system capable of identifying and flagging
suspicious activities.
 To use computer vision to analyze video feeds efficiently and
accurately.
 To design a scalable system that can be integrated with existing
surveillance infrastructure.
 To reduce human dependency in monitoring tasks and enhance the
speed of threat detection.

Dept. of CSE,AITM 4
KEY FEATURES
• Monitoring: Detects activities and motionless objects live.
• Deep Learning Algorithms: LSTM (Long-Short Term Memory).
• Dataset: Merged datasets include Fighting Dataset, Real Life Violence
Situations.

Dept. of CSE,AITM 5
WORKFLOW
1. Preprocessing:
• Frames extracted from videos.
• Data augmentation and normalization.
• Videos classified into normal and suspicious categories.
2. Detection:
• Activity Recognition: LSTM(Long Short Term Memory)
processes sequential video frames.
• Frame Extraction : ViT(Vision Transformer) used to extract the
features from the video frames.

Dept. of CSE,AITM 6
SYSTEM ARCHITECTURE DIAGRAM

Input Frame Resize and


Video Extraction Normalization

Trained
Model

Violence Normal weaponized


CONCLUSION
• Impact:
• Enhances security by automating detection and reducing
operator load.
• Future Scope:
• Expand dataset diversity.
• Optimize algorithms for faster detection in low-resource
environments.

Dept. of CSE,AITM 8
Thank you

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy