Identification of ASL Hand Gesture Language Using Deep Learning Technique
摘要
There are many opportunities for study and modeling in the area of gesture categorization and identification. These are helpful for the community of hearing-impaired persons as well as for the evolution of new vision-based devices. A strong hand gesture recognition system should be created in order to recognize words effectively, as hand gesture interpretation relies heavily on this ability. American Sign Language (ASL) is a complete natural language with syntax that is different from English and linguistic elements that are comparable to spoken languages. In this work, VGG19, a CNN model was employed for gesture classification. The model is evaluated on two datasets and has an accuracy of 99.92% for the RGB dataset, 98.73% for the Depth dataset, and 96.54% for the dataset combining both. For the recognition part camera PyCharm software is used that will help the deaf and mute community to understand and learn easily.