WANT TO ADD DEEPFAKE DETECTION OF VIDEO , USING TRAINED MODEL

[AbheetHacker4278]

Is your feature request related to a problem? Please describe. A clear and concise description of what the problem is. Ex. I'm always frustrated when [...]

Deepfake Videos are major challenge in todays world , so i introduce my Deepfake detection model using Resnext50 and LSTM NN Model using Python(Django), HTML , CSS , JS.

DeepFake Detection using Res-Next50 model which is been trained using various datasets like , Celeb-Df Dataset , FaceForensics++ Dataset , Challenge Dataset , which can do Sequential analyse of the video by dividing the video into 20 parts and analyse those parts on the basis of face tone , color , replaced face structrure(nose , eyes , mouth).

Describe the solution you'd like A clear and concise description of what you want to happen.

Can Easily Tell Which Video is FAKE or REAL

Describe alternatives you've considered A clear and concise description of any alternative solutions or features you've considered.

I Want to add Seperate Section in Navbar for Deepfake detection

Additional context Add any other context or screenshots about the feature request here.

Please Assign this Issue to Me under GSSOC24

[github-actions[bot]]

You've successfully raised your issue, We'll get back to you soon. Don't forget to star⭐ the Repo.

[4darsh-Dev]

⚠️[URGENT]⚠️ Reminder for GSSoC 24' Contributors Please join our CogniGuard Project #general GSSoC'24 Discord channel: Here you can discuss your queries and doubts with your mentors. See #109 issues section.

[4darsh-Dev]

⚠️[URGENT]⚠️ Reminder for GSSoC 24' Contributors Please join our CogniGuard Project #general GSSoC'24 Discord channel: Here you can discuss your queries and doubts with your mentors. See #109 issues section.

[AbheetHacker4278]

@4darsh-Dev please assign this issue to ME under GSSOC24

[4darsh-Dev]

@AbheetHacker4278 you've good idea, but our solution doesn't align with deepfake detection if you want to contribute,

1. we can work on fine-tuning a deep learning model for object detection to find visual based dark pattern using CNN or something.
2. A model also needed using OCR, Text Extraction from images These include changing button sizes, color , or just removing the cross button itself,

   Visual Dark pattern sample

   

[AbheetHacker4278]

@4darsh-Dev , I respect your views , but working on this issues required more time and lots of work required , so better , i can do any other thing for you , like improve your Website Security.

[4darsh-Dev]

sure @AbheetHacker4278 you can look for security bugs in the website. Feel free to share your views on improving the security.