[2604.06220] Development of ML model for triboelectric nanogenerator based sign language detection system

Electrical Engineering and Systems Science > Signal Processing arXiv:2604.06220 (eess) [Submitted on 26 Mar 2026] Title:Development of ML model for triboelectric nanogenerator based sign language detection system Authors:Meshv Patel, Bikash Baro, Sayan Bayan, Mohendra Roy View a PDF of the paper titled Development of ML model for triboelectric nanogenerator based sign language detection system, by Meshv Patel and 3 other authors View PDF HTML (experimental) Abstract:Sign language recognition (SLR) is vital for bridging communication gaps between deaf and hearing communities. Vision-based approaches suffer from occlusion, computational costs, and physical constraints. This work presents a comparison of machine learning (ML) and deep learning models for a custom triboelectric nanogenerator (TENG)-based sensor glove. Utilizing multivariate time-series data from five flex sensors, the study benchmarks traditional ML algorithms, feedforward neural networks, LSTM-based temporal models, and a multi-sensor MFCC CNN-LSTM architecture across 11 sign classes (digits 1-5, letters A-F). The proposed MFCC CNN-LSTM architecture processes frequency-domain features from each sensor through independent convolutional branches before fusion. It achieves 93.33% accuracy and 95.56% precision, a 23-point improvement over the best ML algorithm (Random Forest: 70.38%). Ablation studies reveal 50-timestep windows offer a tradeoff between temporal context and training data volume, yielding 84.13% ac...