Semanur Kumaş Successfully Defended Her Thesis

Semanur Kumaş successfully defended her master's thesis titled “Autonomous Semantic Space Informatics of Mobile Robots” on 22.01.2025.

This thesis is concerned with the problem of autonomous semantic localization in mobile robots.  The importance of this problem stems from the increasing demand for using robots for service tasks.  This requires the robot to build a semantic location map, store this model in its memory and use it when needed.  A semantic map is defined by a labeled graph with edges that encode the nodes corresponding to objects and their relationships.  This thesis addresses this problem by presenting a complete framework for creating and using semantic place maps. In the proposed framework, unlike previous work, objects are represented as point cloud objects, which are detected by tracking candidate and hence class-independent point cloud objects over the incoming RGB-D stream. The tracking is done by considering features such as location, shape, color, key points and neighboring objects. While the robot is tracking point cloud objects, it tries to recognize related objects through object recognition. These objects are then used to progressively build the semantic location map. These maps can then be used in the robot's subsequent visits to places of interest. The experimental evaluation shows that the recognition performance of the semantic place map depends on both object recognition and the routes taken during the first visits of the places.