Seaspray is an end-to-end system designed to counter CAPTCHA cloaking, a technique used by phishers to hide malicious content from automated detection. Building on approaches like Crawl-shing, Seaspray first identifies suspicious sites, detects CAPTCHAs using YOLO, and classifies and solves them with a Vision Transformer to reveal hidden content. It then verifies brand legitimacy via logo detection with YOLO and visual embeddings using a ViT encoder, matched against known brands. By unifying CAPTCHA de-cloaking and brand verification within a Transformer-based framework, Seaspray offers a modular and robust solution.
-
Notifications
You must be signed in to change notification settings - Fork 0
Sapienza University of Rome - Data Network & Security 2024/25/2 - Exam Report
License
muradhuseynov1/Seaspray-Transformers-applied-to-CAPTCHA-cloaked-phishing-websites
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
About
Sapienza University of Rome - Data Network & Security 2024/25/2 - Exam Report
Resources
License
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published