IoT-Driven Optimal Lightweight RetinaNet-Based Object Detection for Visually Impaired People

Abstract

Visual impairment is one of the major problems among people of all age groups across the globe. Visually Impaired Persons (VIPs) require help from others to carry out their day-to-day tasks. Since they experience several problems in their daily lives, technical intervention can help them resolve the challenges. In this background, an automatic object detection tool is the need of the hour to empower VIPs with safe navigation. The recent advances in the Internet of Things (IoT) and Deep Learning (DL) techniques make it possible. The current study proposes IoT-assisted Transient Search Optimization with a Lightweight RetinaNet-based object detection (TSOLWR-ODVIP) model to help VIPs. The primary aim of the presented TSOLWR-ODVIP technique is to identify different objects surrounding VIPs and to convey the information via audio message to them. For data acquisition, IoT devices are used in this study. Then, the Lightweight RetinaNet (LWR) model is applied to detect objects accurately. Next, the TSO algorithm is employed for fine-tuning the hyperparameters involved in the LWR model. Finally, the Long Short-Term Memory (LSTM) model is exploited for classifying objects. The performance of the proposed TSOLWR-ODVIP technique was evaluated using a set of objects, and the results were examined under distinct aspects. The comparison study outcomes confirmed that the TSOLWR-ODVIP model could effectually detect and classify the objects, enhancing the quality of life of VIPs.