Paper Title
DEEP LEARNING BASED SYSTEMFOR SIGN LANGUAGE TRANSLATION

Abstract
Abstract - The main purpose of this project is to realize Human Augmentation in the real-time sign language translation system to improve the quality of life for the dumb. Research and build a real-time sign language translation system suitable for general daily life fields, and an integrated system for retouching sentences. The main work is to establish a real-time sign language translation system suitable for general daily life fields, and an integrated system for retouching sentences. By exploring common dialogues in various fields of daily life, the construction of sign language action video dataset and data augmentation can be completed. Design a CNN+LSTM based sign language recognition network, and conduct model training, performance evaluation and optimization with the dataset collected by the public and this project. Additionally, the method for the sign language translation of concatenating into sentences is proposed. Keywords - Human Augmentation, Deep Learning, Sign Language Recognition, Sign Language Translation.