Sign language is a universally used non-verbal communication method that is commonly adopted by those with hearing and speaking disabilities. Sign language recognition is a topic that has been researched extensively and has different methods to do so. Existing systems use CNN, RNN to perform image recognition and display the corresponding text. Other existing systems used hardware intensive or require special equipment to perform the same task. There exists the need of a method to recognize sign language gestures and produce a system where the recognized gestures are converted to speech. The authors intend to use Convolutional Neural Networks to train the model to perform image recognition and translation of text to speech using a custom Indian Sign Language dataset. This will provide those with disabilities a simple method to converse with those who do not understand sign language.