A Lightweight SLR System based on MobileNetV3, BiLSTM, and Attention Mechanism

Hu, Haiyan (2025) A Lightweight SLR System based on MobileNetV3, BiLSTM, and Attention Mechanism. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (4MB) | Preview
Preview	PDF (Configuration Manual) Download (321kB) | Preview

Abstract

This project seeks to develop a system that recognizes American Sign Language (ASL) word level videos. I begin by using MobileNetV3-large to extract image features before using a BiLSTM(Bidirectional Long Short Term Memory Network) to model temporal information. Furthermore, I also add an attention mechanism to bring better attention to the key frames in the video to improve the recognition. Methods such as image augmentation and label smoothing were used when testing the model to make it more generalizable and stable. In general, the sum of the project was developed and validated based on the WLASL-300 dataset. Ultimately, I got good recognition results while keeping the model structure relatively simple which aligned with the project’s main focus: to adapt an efficient model structure and optimize the training strategy to build a base for further applications in the future.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Raj, Kislay UNSPECIFIED
Uncontrolled Keywords:	Sign Language Recognition; MobileNetV3; BiLSTM; Attention Mechanism; DataAugmentation
Subjects:	Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence > Computer vision Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence > Computer vision P Language and Literature > P Philology. Linguistics > Semiotics > Language. Linguistic theory > Gesture. Sign language
Divisions:	School of Computing > Master of Science in Artificial Intelligence
Depositing User:	Ciara O'Brien
Date Deposited:	28 May 2026 14:12
Last Modified:	28 May 2026 14:12
URI:	https://norma.ncirl.ie/id/eprint/9323

Actions (login required)

View Item