Abstract

Language translator for Indian language to English have been developed and it have proven to a challenging domain due to large combination of character in india scripts such as Tamil, Kannada and Hindi. In this paper we propose a system which captures Indian printed character and translates it into English, we have discussed the various method and machine learning model that was used to build this system with a accuracy of 87%.In addition the paper also includes the project .A Generic sensor Frame for the webOSplatform.A centralized daemon to control and access all the sensors that are connected to the webOS device.

Key words: OCR, Seq2Seq, Tesseract and Sensor Framework