Abstract Sign Language (SL) is the method deaf and mute people use to communicate. In order to have a dialogue with this community of people, one should learn sign language or have an interpreter who can translate the speech both ways. However, both options are difficult. There is no significant importance stressed on hearing people to learn sign language even today. Finding an experienced and well trained interpreter is difficult and costly. Hence, a system which can interpret SL and translate to the local spoken language is desirable. The paper reviews existing SL recognition systems. It presents work done on SL recognition followed by work done in India on Indian Sign Language (ISL). It concludes by proposing usage of artificial neural networks for ISL Recognition