Deaf–mute people interconnect through Indian Sign Language with all communities. Demanding of Sign Language Interpreter is increased day to day for solving gap of communication. Deep CNN plays key role for classification problem. We propose the Deep convolution neural network model for classification of alphabet signs. The model is prepared with addition of six convolutional layers and three fully connected layers. The alphabet signs with 3629 images of A–Z dataset are developed in college laboratory with help of students. Experiment is performed with 1744 sample images for training, 437 sample images for validation and 1448 sample images for testing among 3629 images. Proposed method is achieved 96% model classification accuracy. Training and validation datasets accuracy is achieved with 99% and 94.05% respectively.
CITATION STYLE
Dangarwala, K. J., & Hiran, D. (2020). Alphabet classification of indian sign language with deep learning. In Lecture Notes on Data Engineering and Communications Technologies (Vol. 46, pp. 569–576). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-030-38040-3_64
Mendeley helps you to discover research relevant for your work.