Deepfake Image Detection
using Machine Learning and
Deep Learning
ABSTRACT
In recent years, advances in AI, Machine Learning and Deep Learning
have made it easier to create and edit media like images, videos and audio.
While these tools can be used for good but they are also being misused to
create deepfakes that cause threats to privacy, security and trust in digital
content.
Detection of these deepfakes is necessary in order to reduce threats and
maintain security.
This project aims to create a system that detects deepfakes using Machine
Learning algorithms.
Introduction:
Deepfake is a type of technology that uses AI to create fake but realistic looking videos
,images or audio.
It trains AI on real data like photos , videos or recordings of a person to create new content
that mimics their looks, voice or movements.
They are often being used for harmful reasons , like spreading false information or
impersonating people.
Deepfakes are getting better and harder to spot because they look and sound so real.
LITERATURE SURVEY
Existing System:
Various methods have used for detecting deepfake such as
Support Vector Machine
Random Forest
Face-Tracking Techniques
Logistic Regression
Limitations
Scalability
Lack of granularity
Simplicity
Sensitivity to noisy data
Proposed System:
We propose a Convolutional Neural Network(CNN) based system that overcomes the
limitations of Existing systems .
It makes deepfake detection more robust, adaptable and reliable.
BENIFITS:
Better Handling of High-Dimensional Data
Captures Nonlinear Relationships
Scalability with Big Data
Adapts to New Deepfake Techniques
Software Requirements
IDE: Visual Studio Code or any other python supported IDE
Python: Version 3.6 or higher
Required Python Libraries:
OpenCV
Pandas
Numpy
UML DIAGRAMS
CLASS DIAGRAM
ACTIVITY DIAGRAM
OUTPUTS
HOME PAGE
Thank you