Skip to content

Latest commit

 

History

History
56 lines (37 loc) · 1.71 KB

README.md

File metadata and controls

56 lines (37 loc) · 1.71 KB

CAPTCHA Recognition Using Deep Learning

CAPTCHA Recognition Using Deep Learning is a powerful project designed to solve complex CAPTCHA challenges with advanced AI techniques. It combines a custom-trained YOLO model for character detection and a neural network for recognition, achieving high accuracy in diverse and distorted CAPTCHA formats. The project also includes an API for seamless integration into real-world applications.

Features

  • High Accuracy: Solves CAPTCHAs with over 90% accuracy.
  • Custom YOLO Model: Detects and localizes characters in challenging images.
  • Extensive Dataset: Includes a collection of difficult CAPTCHA images and a personalized dataset for training.
  • API for Deployment: Offers a user-friendly API to integrate CAPTCHA-solving functionality into your projects.
  • Comprehensive Codebase: Fully documented scripts for training, predictions, and API usage.

Getting Started

Prerequisites

  • Python 3.8+
  • TensorFlow 2.x
  • YOLO libraries
  • Install dependencies using: 
     python makeapi.py
    Access the API at http://localhost:8000. (with useapi.py)

Example Results

image

Applications

  • Security Testing: Identify weaknesses in CAPTCHA implementations.
  • Automation: Streamline tasks requiring CAPTCHA solving.

License

This project is licensed under the MIT License.

Contributing

Contributions are welcome! Feel free to open issues or submit pull requests.

Contact

For questions or feedback, reach out to Taha.