Publication: Progression Learning Convolution Neural Model-Based Sign Language Recognition Using Wearable Glove Devices
0
0
Issued Date
2024-04-01
Resource Type
eISSN
20793197
Scopus ID
2-s2.0-85191354959
Journal Title
Computation
Volume
12
Issue
4
Rights Holder(s)
SCOPUS
Bibliographic Citation
Computation Vol.12 No.4 (2024)
Suggested Citation
Liang Y., Jettanasen C., Chiradeja P. Progression Learning Convolution Neural Model-Based Sign Language Recognition Using Wearable Glove Devices. Computation Vol.12 No.4 (2024). doi:10.3390/computation12040072 Retrieved from: https://hdl.handle.net/20.500.14740/20445
Author(s)
Corresponding Author(s)
Other Contributor(s)
Abstract
Communication among hard-of-hearing individuals presents challenges, and to facilitate communication, sign language is preferred. Many people in the deaf and hard-of-hearing communities struggle to understand sign language due to their lack of sign-mode knowledge. Contemporary researchers utilize glove and vision-based approaches to capture hand movement and analyze communication; most researchers use vision-based techniques to identify disabled people’s communication because the glove-based approach causes individuals to feel uncomfortable. However, the glove solution successfully identifies motion and hand dexterity, even though it only recognizes the numbers, words, and letters being communicated, failing to identify sentences. Therefore, artificial intelligence (AI) is integrated with the sign language prediction system to identify disabled people’s sentence-based communication. Here, wearable glove-related sign language information is utilized to analyze the recognition system’s efficiency. The collected inputs are processed using progression learning deep convolutional neural networks (PLD-CNNs). The technique known as progression learning processes sentences by dividing them into words, creating a training dataset. The model assists in efforts to understand sign language sentences. A memetic optimization algorithm is used to calibrate network performance, minimizing recognition optimization problems. This process maximizes convergence speed and reduces translation difficulties, enhancing the overall learning process. The created system is developed using the MATLAB (R2021b) tool, and its proficiency is evaluated using performance metrics. The experimental findings illustrate that the proposed system works by recognizing sign language movements with excellent precision, recall, accuracy, and F1 scores, rendering it a powerful tool in the detection of gestures in general and sign-based sentences in particular.