MACHINE LEARNING ALGORITHMS FOR ASL IMAGE RECOGNITION WITH LENET5 FEATURE EXTRACTION
Main Article Content
Abstract
American Sign Language is used by mute and deaf people so that they can interact with the people around them. It is used by approximately 2,50,000-5,00,000 Americans (and some Canadians) of all ages. Over the period of time, many have proposed different methods for recognition of ASL. Sign language Recognition is a complex technical problem due to the difficulty of visual analysis of hand motions and the highly structured nature of sign language. Hence the accuracy is not achieved. To enhance this accuracy, the proposed system compares different machine learning classification algorithms using Lenet5 architecture for feature extraction. Lenet5 is one of the architectures of Convolution Neural Network (CNN). The proposed system uses machine learning algorithms like Neural Network, Decision Tree Classifier, K-Nearest Neighbour and Support Vector Machine. There is a raise in accuracy using Neural Network algorithm when compared to other machine learning algorithms. The proposed solution was tested on data samples from ASL data sets and achieved an overall accuracy of 99.99% using Neural Network.
Downloads
Article Details
COPYRIGHT
Submission of a manuscript implies: that the work described has not been published before, that it is not under consideration for publication elsewhere; that if and when the manuscript is accepted for publication, the authors agree to automatic transfer of the copyright to the publisher.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgment of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgment of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work
- The journal allows the author(s) to retain publishing rights without restrictions.
- The journal allows the author(s) to hold the copyright without restrictions.