Individuals who are deaf and rely on gesture language face communication challenges as their primary mode of communication is sign language. Sign language movements convey ideas and emotions in a non-verbal manner. It can be difficult to communicate with someone who has hearing loss because people who can hear and talk may have trouble understanding the gestures used in sign language by people who are deaf and mute. Therefore, it is essential to develop systems that can recognise various sign language motions and convey them to a wider audience. Regular people may find it difficult to understand the facial gestures used in sign language, but professionals are needed for medical and legal consultations as well as educational and training programmes. Demand for these services has recently increased, and new ones have also developed, notably video-based remote human interpretation. However, these services have limitations in terms of simple gesture recognition interpretation services, which require a high-speed internet connection. By identifying the user's finger placements, machine learning techniques may be used to study the user's hand movements. Many researchers have suggested numerous algorithms in the last few years that have considerably profited from deep learning approaches. This work will concentrate on investigating deep learning-based vision-based models for gesture identification.