Yarı denetimli eğitilmiş katlamalı filtreler ile görsel nesne takibi

Visual object tracking is a challenging computer vision problem with numerous real-world applications. During recent years, correlation filters have produced excellent results in terms of accuracy and performance for visual tracking problem. Feature properties which are utilized by these filters plays important role on the performance. Recently, deep learning based methods have emerged to learn best features for correlation filters, which have shown promising results. In this thesis, the impact of semi supervised trained convolutional filters for the visual tracking problem will be investigated in order to obtain robust features predicting the object location with high accuracy and being invariant to any kind of apperance change.