Sign language recognition has become a critical research in the field of computer vision as the need of disability solutions grow. Sign language acts as a bridge to reduce the communication gap between normal people and deaf and dumb people. Current sign language identification systems, on the other hand, lack essential characteristics such as accessibility and cost, which are critical for people with speech disabilities to interact with their daily settings. The successful attractive solution is to initiate the sign languages in terms of words and common expressions for daily activities. This will interact the deaf and dumb people by connecting to the outside world more quickly and easily. The sign gestures obtained are processed through popular machine learning and deep learning models for classification accuracy. This chapter discusses the word sign recognition, image processing algorithms for separating the signs from the background, machine learning algorithms, and the complete model set up for sign recognition.