Sign Language Detection Using CNN-YOLOv8l

Sign language is commonly used by speech and hearing-disabled people as a way of exchanging thoughts, emotions, feelings, and needs. There are multiple sign languages used all around the globe. Understanding sign language can be a difficult task for a normal person. At present, speech and hearing-disabled people rely on human translators to make this task easier, but it is not always feasible to have a human translator. Multiple sign language detection models were proposed in the current time, but they are mostly based on old object detection techniques, thus making them slower to detect and less accurate. The objective of the present work is to design and implement a model which detects real-time sign language using deep learning. The proposed system uses a YOLOv8 model to train the dataset. A deep learning neural network is like the human brain which uses a combination of different factors like inputs, weights, and biases to do different tasks, such as identifying and classifying objects. Deep learning algorithms perform better when they are trained with a large amount of data. The performance of the proposed model has been evaluated based on accuracy, precision, and recall. The dataset consists of Indian sign Language having9991 images of 26 alphabets and 9 numeric digits. The proposed model shows a training accuracy of 98.9% for 20 epochs of training.