Cilt:64 Sayı:01 (2022)

Permanent URI for this collection

https://dspace.ankara.edu.tr/handle/20.500.12575/86465

Browse

Browsing Cilt:64 Sayı:01 (2022) by Author "Bilgisayar Mühendisliği"

Now showing 1 - 2 of 2
  • No Thumbnail Available
    Item
    A method for analyzing suspect-filler similarity using convolutional neural networks
    (Ankara Üniversitesi Mühendislik Fakültesi, 2022) Ar, Yılmaz; Bilgisayar Mühendisliği; Mühendislik Fakültesi
    Eyewitness misidentifications are one of the leading factors in wrongful convictions. This study focuses on the structure of the lineups, which is one of the factors that cause misidentification, and the use of artificial intelligence (AI) technologies in the selection of fillers to be included in the lineups. In the study, AI-based face recognition systems are used to determine the level of similarity of fillers to the suspect. Using two different face recognition models with a Convolutional Neural Network (CNN) structure, similarity threshold values close to human performance were calculated (VGG Face and Cosine similarity = 0.383, FaceNet and Euclidean l2 = 1.16). In the second part of the study, the problems that are likely to be caused by facial recognition systems used in the selection of fillers are examined. The results of the study reveal that models responsible for facial recognition may not suffice alone in the selection of fillers and, an advanced structure using CNN models trained to recognize other attributes (race, gender, age, etc.) associated with similarity along with face recognition models would produce more accurate results. In the last part of the study, a Line-up application that can analyze attributes such as facial similarity, race, gender, age, and facial expression, is introduced.
  • No Thumbnail Available
    Item
    Classification of five different rice seeds grown in Turkey with deep learning methods
    (Ankara Üniversitesi Mühendislik Fakültesi, 2022) Tuğrul, Bülent; Bilgisayar Mühendisliği; Mühendislik Fakültesi
    The increase in the world population and harmful environmental factors such as global warming necessitate a change in agricultural practices with the traditional method. Precision agriculture solutions offer many innovations to meet this increasing need. Using healthy, suitable and high-quality seeds is the first option that comes to mind in order to harvest more products from the fields. Seed classification is carried out in a labor-intensive manner. Due to the nature of this process, it is error-prone and also requires a high budget and time. The use of state-of-the-art methods such as Deep Learning in computer vision solutions enables the development of different applications in many areas. Rice is the most widely used grain worldwide after wheat and barley. This study aims to classify five different rice species grown in Turkey using four different Convolutional Neural Network (CNN) architectures. First, a new rice image dataset of five different species was created. Then, known and widely applied CNN architectures such as Visual Geometry Group (VGG), Residual Network (ResNet) and EfficientNets were trained and results were obtained. In addition, a new CNN architecture was designed and the results were compared with the other three architectures. The results showed that the VGG architecture generated the best accuracy value of 97%.