American-Sign-Langugae

The application is developed using Python3, OpenCV and with concepts of Neural Networks it is trained on DarkNet 53 and for the real time object detection I am using Yolo v3. It detects the 1 hand representation of number 1-10 which will be pointed by the user in real time using American Sign Language

Please consider following this project's author, Abdullah Mujahid, and consider starring the project to show your :heart: and support.

For live demo

For Dataset if I don't get back to you Dataset & Model

Project Features.

• Detects real time object and recognize which digit it is from 1 - 10 .
• Detects number from a video from 1 - 10 digits.

Prerequisites

• Python3
• OpenCV
• Numpy
• Tkinter
• Pillow

Yolo v3

• For Real time object Detection Yolo (You Only Look Once) is used.

This application is tested on Window Linux and macOS*

• Clone this Repo

        ∧∧
　　 ( ´◔ω◔`)　＜・・・・・・ $ git clone https://github.com/abdullahmujahidali/American-Sign-Language.git
　 ＿|　⊃／(＿＿_
／　└-(＿＿＿_／
￣￣￣￣￣￣￣

　＜⌒／ヽ-､＿＿ 　 　　*thanks
／＜_/＿＿＿＿／

• Install requirements

$ pip3 install -r requirements.txt

Email @ [email protected] and i'll send you weights and cfg file for American Sign Language

• Run

$ pyhton3 main.py

make sure before running code paste .cfg and .weights files

Author

Abdullah Mujahid

• LinkedIn Profile
• GitHub Profile
• Twitter Profile

License

Copyright © 2020, Abdullah Mujahid. Released under the MIT License.