Date

2021-9-7

Author

Köksal, Aybora

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

670
views

2189
downloads

Cite This

In this thesis, single target object detection problem is examined. Object detection is a problem that aims defining all of the objects of interest with their pre-defined classes in an image, or in a series of images. The main objective of this thesis is to exploit spatio-temporal information for performance enhancement during moving object detection. To this extent, modern object detection algorithms which are based on CNN architectures are analyzed. Based on this analysis, state-of-the-art techniques which are focused on utilization of temporal information on object detection are studied and some new methods are proposed. Apart from the aforementioned analysis, some additional studies are also covered. The state-of-the-art object detection algorithms are based on the deep neural networks which are trained via supervised learning. Since these methods need lots of annotated data which also requires a lot of human labor, automatic and semi-automatic annotation methods are employed to overcome this problem. However, automated annotation methods sometimes result in erroneous annotations and effects of these type of errors are examined. A novel method is proposed to correct some type of such annotation errors. This effort is extended with another preceding work, which suggests an alternative method for semi-automatic bounding box annotation for object detection with a significant reduction on annotation effort.

Subject Keywords

Object detection, Video object detection, Tracking with detection, Annotation errors, Annotation correction, Semi-automatic annotation

URI

https://hdl.handle.net/11511/92140

Collections

Graduate School of Natural and Applied Sciences, Thesis