SignPE: Real-Time Peruvian Sign Language Recognition through a MediaPipe-LSTM Architecture

Rojas D.C.G.; Morales J.J.C.; Silva P.D.R. · Proceedings - 2025 5th International Conference on Information Systems and Software Technologies, ICI2ST 2025 (2025) · DOI: 10.1109/ici2st68465.2025.00025

Resumen

More than half a million people in Peru with hearing loss face significant barriers to learning Peruvian Sign Language (PSL) due to a lack of accessible resources. This article presents SignPE, an intelligent system for real-time PSL sign recognition. The methodology uses MediaPipe to extract gesture features, which are then processed by a bidirectional LSTM neural network model. This research represents a significant step toward developing an effective, scalable, and inclusive web platform that contributes to the communicative inclusion of the deaf community in Peru.

Instituciones: Universidad Peruana de Ciencias Aplicadas