Sign Explainer An Explainable AI Enabled Framework for Sign Language Recognition With Ensemble Learning

Aim:

         To enhance the accuracy, robustness and interpretability of sign language recognition system

Synopsis:

         Sign Language recognition is a pioneering framework designed to advance the field of Sign Language Recognition (SLR) through the innovative application of Ensemble Deep learning models. The primary goal of this research is to significantly improve the accuracy, resilience and interpretability of SLR systems. Leveraging the unique features of ResNet within an ensemble learning paradigm. The key component of InceptionResNetv2 architecture is its deep and effective feature extraction capabilities. The utilization of Inception ResNet model enhances the model ability to capture intricate details crucial for accurate sign language recognition. This framework is also to scale seamlessly, accommodating an expanding vocabulary of signs, diverse users and dynamic environmental conditions without compromising performance

Proposed System:

          Sign language recognition is a critical aspects of facilitating communication for the deaf and hard-of-hearing community. This research introduces an innovative approach utilizing the inceptionResNetV2 architecture for robust and accurate sign language classification. A diverse dataset is employed, encompassing a variety of sign language gestures captured from real world scenarios. The proposed methodology involves the transfer learning technique, adapting the InceptionResNetV2 model to the specific requirements of sign language recognition. The model is fine-tuned with a new classification layer to suit the unique characteristic of the dataset. Additionally, data augmentation techniques are applied to enhance the models generalization capability. Training and Validation are conducted on a carefully partitioned dataset and the performance is evaluated through a comprehensive set of metrics. The resulting model demonstrates high accuracy, showcasing its potential for practical applications in real time sign language recognition scenarios. The study contributes to the ongoing efforts in developing efficient human-machine communication system for the deaf and hard of hearing community. The InceptionResNetV2 based model offers promising results, emphasizing its efficiency in advancing the state of the art in sign language recognition technology.