| Autor: |
Das, Haritha V., Mohan, Kavya, Paul, Linta, Kumaresan, Sneha, Nair, Chitra S. |
| Zdroj: |
Multimedia Tools & Applications; Feb2024, Vol. 83 Issue 5, p13543-13555, 13p |
| Abstrakt: |
One of the major challenges faced by medical practitioners is to facilitate hassle free communication with speech impaired people. It is difficult for speech impaired people to communicate their illness to a physician as most physicians do not understand sign language. Sign languages are languages that use the visual-manual modality to convey meaning. Sign languages are expressed through manual articulations in combination with non-manual elements. Indian Sign Language (ISL) is the most widely used sign language by the deaf and hearing-impaired community in India. Hearing impaired people face challenges during consultation with doctors in communicating the symptoms effectively due to the inability of doctors to understand sign language. The situation becomes even worse when the consultation is through an online platform. The existing models recognise hand gestures from videos, compare the recognised hand spatial orientation with the annotated dataset, and identify the letters or words. The drawbacks of the existing systems are, they neither include the entire vocabulary in the dataset nor are they contextual based. Hence these models are not very effective in a real-life communication situation. The datasets available are limited to recognition of alphabets, digits, some words like greeting terms, bank transaction terms etc. There has been little or no work done related to recognition of medical symptoms from Indian Sign Language. The proposed model will recognise patients' hand gestures in the video captured and translate it to corresponding words. Short sentences will be framed from the words recognised, in order to ensure clear communication. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
Full text from SpringerLink