Published Paper Details:

Sign language is a visual language that is mainly used by deaf people and the Hard-of-Hearing community as their primary language. Unlike acoustically expressed sound patterns, sign language uses hand signals, gestures, facial expressions, and body language. Sign language may also help the people with disabilities, namely Autism, Cerebral Palsy, Apraxia of speech and Down Syndrome. The objective of our project is to reduce the communication gap between normal people and disabled people by providing a virtual animation translator for sign language. Speech to sign translator is a user-friendly web application that converts speech/text to sign language animation and also reduces the amount of effort spent on communication. The system consists of both speech and text conversion using natural language processing (NLP) and machine learning techniques. Input speech is captured through a microphone then it is translated to text through Speech recognition using Mel-frequency cepstral coefficients (MFCC) features. The recognized text preprocessed using NLTK and resultant words looks for a word match in ASL dataset. The database contains a certain number of pre-recorded video animation signs using 3D avatar where mainly there is one video clip per each basic word. If a match occurred for a word, then the equivalent ASL translation will be displayed corresponding to that word. Otherwise, the word will be finger spelled.