Hand-Sign Detection and Prediction

Table of Content:

1. Cloning the Repository

2. Installation of Our Requirements

3. Testing Model (Optional)

4. Prediction

5. Results:

File Structure:

D:.
├───data
│   ├───test
│   │   ├───0
│   │   ├───1
│   │   ├───2
│   │   ├───3
│   │   ├───4
│   │   └───5
│   └───train
│       ├───0
│       ├───1
│       ├───2
│       ├───3
│       ├───4
│       └───5
├───data_collection
│   └───__pycache__
├───data_modeling
│   ├───testing
│   │   └───__pycache__
│   ├───training
│   │   └───__pycache__
│   └───__pycache__
├───model
├───outputs
└───prediction
    └───__pycache__

1. Cloning the Repository:

You can clone this repository in your directory by entering the command:

$ git clone <repo> <directory>

In our case use this command:

$ git clone git://github.com/itsvivekghosh/Hand-Sign_Detection_and_Prediction.git

2. Installation of Our Requirements:

All the requirement libraries are added into requirements.txt file:

Your can install all the libraries by just entering the below command into your bash or command prompt:

• For Python < 3:

pip install -r requirements.txt

• For Python >= 3:

pip3 install -r requirements.txt

3. Testing The Model: (Optional)

Its very easy to test the model. But, we need to follow some steps to that:

• Open main.py file which is the main directory.
• After Opening this file, uncomment line 27 where the line is testModel(...) .
• Now, run the main.py file for testing the model is working properly or not.

None:- After successfull testing of the Model, the camera will Open immediately next step (Prediction) for solving this, uncomment the line 37 and then go for the further steps..

4. Prediction:

After Successfull Testing (if implemented last step) of the Model, run main.py file using this simple command:

• For Python < 3:

python main.py

• For Python >= 3:

python3 main.py

Note: Please make sure that main.py file has line 37 uncommented out otherwise will generate errors.

5. Results:

After The Predictions from WebCam, the result can be something like these:

About Author:

Code Written by:

Vivek Kumar Ghosh
Uttaranchal University, Dehradun
Bachelor of Technology, Computer Science
E-mail: [email protected]
LinkedIn: https://www.linkedin.com/in/itsvivekghosh

For any bugs and queries feel free to contact the above profiles. Well, we are working on fixing other Bugs.

Happy Coding!