Visual object detection and tracking using local convolutional context features and recurrent neural networks

---

Tezin Türü: Yüksek Lisans

Tezin Yürütüldüğü Kurum: Orta Doğu Teknik Üniversitesi, Mühendislik Fakültesi, Elektrik ve Elektronik Mühendisliği Bölümü, Türkiye

Tezin Onay Tarihi: 2018

Öğrenci: EMRE CAN KAYA

Danışman: ABDULLAH AYDIN ALATAN

Özet:

Visual object detection and tracking are two major problems in computer vision which have important real-life application areas. During the last decade, Convolutional Neural Networks (CNNs) have received significant attention and outperformed methods that rely on handcrafted representations in both detection and tracking. On the other hand, Recurrent Neural Networks (RNNs) are commonly preferred for modeling sequential data such as video sequences. A novel convolutional context feature extension is introduced to a proposal-based detection scheme for improving object detection performance. A comprehensive experimental study is conducted to demonstrate the effectiveness of this newly proposed approach. On the tracking side, the effect of several design choices is investigated for an RNN-based tracking algorithm by the help of comparative experiments. Finally, the proposed context feature based method is combined with the RNN-based tracking framework and a joint detection-tracking framework that outperforms the baseline model is proposed.