UZBEK SIGN LANGUAGE CLASSIFIER BASED ON MACHINE LEARNING

The "Uzbek Sign Language Classifier Based on Machine Learning" presents a groundbreaking approach to enhancing communication accessibility for the deaf and hard-of-hearing community in Uzbekistan. This research focuses on developing a robust machine learning model to recognize and classify Uzbek Sign Language (UzSL) signs, enabling real-time translation and fostering better integration of sign language users into various aspects of society. Utilizing advanced deep learning techniques, particularly Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs), this study addresses the intricate challenges associated with sign language recognition, such as gesture segmentation, feature extraction, and accurate classification. The dataset for this project comprises a comprehensive collection of images and videos of Uzbek sign language gestures. These data points undergo meticulous preprocessing, including normalization and augmentation, to enhance model training. The proposed model architecture leverages CNNs for spatial feature extraction and RNNs for capturing temporal dependencies, ensuring high accuracy in recognizing dynamic sign sequences. Additionally, the integration of transfer learning techniques, employing pre-trained models, significantly improves the model's performance, particularly given the relatively limited size of the dataset. Extensive experimentation and validation are conducted to fine-tune the model's hyper parameters, optimizing its accuracy and robustness. The results demonstrate a promising accuracy rate, highlighting the model's capability to accurately recognize and classify a wide range of Uzbek sign language gestures. Furthermore, this research underscores the importance of developing localized sign language recognition systems, tailored to the specific linguistic and cultural nuances of the target community.

Hali tarjima qilinmagan