Sign language detection website can recognition gesture through webcam in real-time and by uploading images using CNN, image-classification and image-recognition.
Install the environment
conda create -n tf python==3.7.9
conda activate tfInstalling dependencies
pip install -r requirements.txtRunning the django server
python manage.py runserverLink to access the website
127.0.0.1:8000/This dataset has a total of 52,000 images of signs(alphabets); the dataset is separated into two sections: training and testing. The testing dataset has a capacity of 6,500 images, and the training dataset has a total of 45,000 images. In both the training and testing dataset, the images are being classified into 26 categories, each category images labelled with alphabets from A to Z. They are 26 labelled images folders in both the training and testing dataset. In each alphabet labelled folder, they are 250 images of that alphabet in the testing dataset and 1,750 images of that alphabet in the training dataset.
Home Page
Upload Page
Detection Page
Results
