PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
March 21, 2026Applied Sciences2 citationsOpen Access

Dual-Stream BiLSTM–Transformer Architecture for Real-Time Two-Handed Dynamic Sign Language Gesture Recognition

EAEnachi AndreiTCTurcu Corneliu-OctavianCGCulea George

Key Points

  • To develop a novel architecture for recognizing two-handed dynamic gestures in sign language, focusing on independent hand tracking and motion patterns.
  • Proposed a dual-stream BiLSTM–Transformer model architecture with separate encoders for each hand.
  • Utilized MediaPipe Hands for extracting hand landmarks as a preprocessing step.
  • Implemented an attention-based cross-hand fusion mechanism to capture spatial and temporal dependencies.
  • Achieved significant improvements in recognition accuracy for asymmetric gestures.
  • Demonstrated consistent performance gains for synchronized two-handed gestures compared to single-handed baselines.
  • Outlined an efficient solution suitable for real-time sign language recognition.

Abstract

Two-handed dynamic gesture recognition represents a fundamental component of sign language interpretation involving the modeling of temporal dependencies and inter-hand coordination. In this task, a major challenge is modeling asymmetric motion patterns, as well as bidirectional and long-range temporal dependencies. Most existing frameworks rely on early fusion strategies that merge joints, keypoints or landmarks from both hands in early processing stages, primarily to reduce model complexity and enforce a unified representation. In this work, a novel dual-stream BiLSTM–Transformer model architecture is proposed for two-handed dynamic sign language recognition, where parallel encoders process the trajectories of each hand independently. To capture spatial and temporal dependencies for each hand, an attention-based cross-hand fusion mechanism is employed, with hand landmarks extracted by the MediaPipe Hands framework as a preprocessing step to enable real-time CPU-based inference. Experimental evaluation conducted on custom Romanian Sign Language dynamic gesture datasets indicates that the proposed dual-stream-based system outperforms single-handed baselines, achieving improvements in high recognition accuracy for asymmetric gestures and consistent performance gains for synchronized two-handed gestures. The proposed architecture represents an efficient and lightweight solution suitable for real-time sign language recognition and interpretation.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Andrei et al. (2026) studied this question.

synapsesocial.com/papers/69be37aa6e48c4981c6776b2https://doi.org/10.3390/app16062912
Ask AI
Helpful
Bookmark
Share
View Full Paper