Scopus Indexed Paper

Paper Details

A Simple and Mighty Arrowhead Detection Technique of Bangla Sign Language Characters with CNN
Sign Language is argued as the first Language for hearing impaired people. It is the most physical and obvious way for the deaf and dumb people who have speech and hearing problems to convey themselves and general people. So, an interpreter is wanted whereas a general people needs to communicate with a deaf and dumb person. In respect to Bangladesh, 2.4 million people uses sign language but the works are extremely few for Bangladeshi Sign Language (BdSL). In this paper, we attempt to represent a BdSL recognition model which are constructed using of 50 sets of hand sign images. Bangla Sign alphabets are identified by resolving its shape and assimilating its structures that abstract each sign. In proposed model, we used multi-layered Convolutional Neural Network (CNN). CNNs are able to automate the method of structure formulation. Finally the model gained 92% accuracy on our dataset.
Bangla Sign Language, NLP, Computer vision, Machine learning, Image processing, Sign language characters, BdSL, BSL, CNN, Pattern recognition
Md. Sanzidul Islam, Sadia Sultana Sharmin Mousumi, AKM Shahariar Azad Rabby, Syed Akhter Hossain
Journal or Conference Name
Communications in Computer and Information Science
Publish Year