Deep Learning-based Multiple Objects Detection and Tracking System for Socially Aware Mobile Robot Navigation Framework

Multiple objects (including humans) detection and tracking system plays an essential role in socially aware mobile robot navigation framework. Because, it provides an important input for the remaining modules of the framework. In this paper, we propose an efficient multiple objects detection and tracking system for mobile service robots in dynamic social environments using deep learning techniques. The proposed system consists of two steps: (1) multiple objects detection, and (2) multiple objects tracking. In the first step, the RGB image-based multiple objects detection is made use of to detect objects in the mobile robot's vicinity using a convolutional neural network. In the second stage of system, the detected objects are tracked using a deep simple online and realtime tracking technique. The experimental results indicate that, the proposed system is capable of detecting and tracking multiple objects including humans, providing significant information for the socially aware mobile robot navigation framework.