Machine Learning Techniques for Urdu Audio Feedback for Visual Assistance: A Systematic Literature Review


Article PDF :

Veiw Full Text PDF

Article type :

1

Author :

M. Hanif,T. Ahmad,M. Aslam,M. Waseem

Volume :

60

Issue :

2

Abstract :

Abstract Visually impaired individual faces many challenges when comes to object recognition and routing inside or out. Despite the availability of numerous visual assistance systems, the majority of these system depends on English auditory feedback, which is not effective for the Pakistani population, since a vast population of Pakistanis cannot comprehend the English language. The primary object of this study is to consolidate the present research related to the use of Urdu auditory feedback for currency and Urdu text detection to assist a visually impaired individual in Pakistan. The study conducted a comprehensive search of six digital libraries, resulting in 50 relevant articles published in the past five years. Based on the results, a taxonomy of visual assistance was developed, and general recommendations and potential research directions were provided. The study utilized firm inclusion/exclusion criteria and appropriate quality assessment methods to minimize potential biases. Results indicate that while most research in this area focuses on navigation assistance through voice audio feedback in English, the majority of the Pakistani population does not understand the language rendering such systems inefficient. Future research should prioritize object localization and tracking with Urdu auditory feedback to improve navigation assistance for visually impaired individuals in Pakistan. The study concludes that addressing the language barrier is crucial in developing effective visual assistance systems for the visually impaired in Pakistan.

Keyword :

N/A