Kyambogo University Lecturer Nkalubo Leonard from the Department of Computer Science is researching on how to “Integrating Computer Vision and Natural Language Processing” to guide blind movements.
He confirmed that his research is on a case study based on the University’s disability unit (UNISE) and health informatics.
The research was revealed during the 2019 Federated Conference on Computer Science and Information Systems – Doctoral Symposium (FedCSIS) (DS-RAIT’19) which was held in Lepzig, Germany from 1st – 4th September, where Mr Nkalubo got a chance to present his research ideas in a paper.
Machine learning concepts in artificial intelligence such as computer vision and Natural Language Processing (NLP) have become so popular in current trends of research in Computer Science and Health Informatics.
Mr. Nkalubo believes that the integration of these two concepts can help solve many problems in medical health and disability problems which can be done through generating a description of medical images which can be used by doctors and making a system which sees the surrounding and gives a spoken description of the same can be used by blind people.
Vision impairment is one of the most problems faced by the elderly, implementing such a solution will give a higher advantage to reducing the hindrances faced by the visionary impaired persons.
The implementation requires an advanced Smartphone equipped with language intelligence capabilities to be attached to the blind person to capture surrounding images and is then connected to a central server equipped with a faster region convolutional neural network algorithm image detection algorithm to recognize images and multiple obstacles.
The server sends the results back to the Smartphone which are then converted into speech for the blind person’s guidance. The abstract revealed.
The research which has already been published on //annals-csis.org/ marketed Kyambogo University as a resource centre to provide excellent scholars who can share their research knowledge in advanced academic forums in Europe and it will increase Kyambogo on global University rankings.
The research area is believed to provide students with blind disabilities by offering quick, accurate and user-friendly machine learning solutions and also strengthen the graduate programs in the Department of Computer Science of Kyambogo.
Nkalubo also thinks that the concepts and tools in the research paper can also be enlarged to making systems which can convert spoken content in form of some image which may help persons who do not own ability of speaking and hearing.
“The research is going on and we have already simulated the experiment using the Makerere AI lab where there is a GPU and faster AI WIFI but I am going to get around 30M from the staff development, I will run the research results in UNISE.” He told Cyclone Times.
“By the end of October, there will be another publication with results in print. But already preliminary results are out and very positive. I will need to buy that Google 3xL phone and a GPU for use herein Kyambogo Univerity (KYU), then we shall use any fast WIFI.” Leonard added.
The paper is also expected to be published in the IEEE digital library this week.
Original Source: //www.cyclonetimes.com/kyambogo-lecturer-moves-to-integrate-computer-vision-natural-language-processing-to-guide-blind-movements/