Scribd
Upload
42 views12 pages

Deepfake Image Detection: SANDEEP KUMAR (2203555) YADVEER (2203578) SUNNY DUTT (2203567)

The document discusses a project focused on developing a system to detect deepfake images using advanced AI techniques such as Convolutional Neural Networks (CNNs) and Generative Adversarial Networks (GANs). It outlines objectives including high accuracy detection, comparison with other deep learning models, and optimization for real-time use. The project also emphasizes the importance of preventing misinformation and maintaining media integrity through technical and economic feasibility studies.

Uploaded by

sanatani0mantra
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
42 views12 pages

Deepfake Image Detection: SANDEEP KUMAR (2203555) YADVEER (2203578) SUNNY DUTT (2203567)

The document discusses a project focused on developing a system to detect deepfake images using advanced AI techniques such as Convolutional Neural Networks (CNNs) and Generative Adversarial Networks (GANs). It outlines objectives including high accuracy detection, comparison with other deep learning models, and optimization for real-time use. The project also emphasizes the importance of preventing misinformation and maintaining media integrity through technical and economic feasibility studies.

Uploaded by

sanatani0mantra
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

DEEPFAKE IMAGE

DETECTION

• SANDEEP KUMAR (2203555)


• YADVEER ( 2203578 )
• SUNNY DUTT (2203567 )
DEEPFAKE IMAGE
DETECTION

Deepfakes are digital media - video, audio and images,


edited and manipulated using artificial intelligence
What is deepfake? A.I.). This synthetic media content is referred to as the
Deepfakes.
INTRODUCTION
DEEPFAKE TECHNOLOGY USES AI TO CREATE FAKE MEDIA BY
ALTERING IMAGES, VIDEOS, OR AUDIO. WHILE IT HAS POSITIVE
USES IN ENTERTAINMENT AND EDUCATION, IT ALSO CREATES RISKS
LIKE MISINFORMATION, IDENTITY THEFT, AND MANIPULATION.
THIS PROJECT AIMS TO BUILD A SYSTEM TO DETECT DEEPFAKE
IMAGES USING ADVANCED AI TECHNIQUES LIKE CONVOLUTIONAL
NEURAL NETWORKS (CNNS) AND GENERATIVE ADVERSARIAL
NETWORKS (GANS).
OBJECTIVES
1.Detect Deepfake Images: Build a model that can identify fake images
with high accuracy.
2. To Compare with other models of deep learning on the dataset like
ResNet.
3. To optimize the model’s efficiency for real-time deepfake detection with
minimal computational resources.
LITERATURE REVIEW
• ViT for Deepfake Detection: Vision Transformers can
detect fake images, with up to 95% accuracy.
• Artifact Analysis: Detecting lighting inconsistencies
can help identify fakes.
• Real-Time Detection: Lightweight models allow for
quick, real-time detection.
FEASIBILITY STUDY
• Technical Feasibility: Tools like TensorFlow, PyTorch,
and datasets like Kaggles make the project possible.
• Economic Feasibility: The project is cost-effective as it
uses open-source tools.
• Social Impact: This project helps prevent
misinformation and maintains media integrity.
METHODOLOGY
1. Data Collection: Images labeled as ‘real’ or ‘fake’.
2. Model Design: Build a model using ViT.
3. Training: Train the model and validate its performance.
4. Testing: Test the model with new data.
5. Interface: Develop an easy-to-use interface with Flask
TOOLS & FACILITIES
• Software: VS code,Python, TensorFlow, PyTorch, OpenCV, Flask.
• Datasets: Kaggles,Celeb-DF, FaceForensics++.
• Tools for Analysis: Jupyter Notebook for testing and analysis
EXPECTED OUTCOMES
• A model that can accurately detect deepfake images.
• A tool for users to upload and verify images.
• A comparison of our model’s performance with other
existing methods
Output

1.First sample of a Real image :


2.Second sample of a Fake image :
REFERENCES
1. A. Rössler et al., *FaceForensics++: Learning to Detect
Manipulated Facial Images*, IEEE ICCV, 2019.
2. Y. Li et al., *Celeb-DF: A Large-Scale Dataset for DeepFake
Forensics*, IEEE CVPR, 2020.
3. H. H. Nguyen et al., *DeepFake Detection Using CNNs*,
Journal of Digital Forensics, 2021.

You might also like