Bilgisayar Mühendisliği Bölümü Koleksiyonu
Permanent URI for this collectionhttps://hdl.handle.net/20.500.11779/1940
Browse
3 results
Search Results
Conference Object Citation - WoS: 1Citation - Scopus: 2Turcoins: Turkish Republic Coin Dataset(IEEE, 2021) Gökberk, Berk; Gökberk, Berk; Temiz, Hüseyin; 02.02. Department of Computer Engineering; 02. Faculty of Engineering; 01. MEF UniversityIn this paper, we present a novel and comprehensive dataset which contains Turkish Republic coins minted since 1924 and present a deep learning based system that can automatically classify coins. The proposed dataset consists of 11080 coin images from 138 different classes. To classify coins, we utilize a pre-trained neural network (ResNet50) which is pre-trained on ImageNet. We train the pre-trained neural networks on our dataset by transfer learning. The imbalanced nature of the dataset causes the classifier to show lower performance in classes with fewer samples. To alleviate the imbalance problem, we propose a StyleGAN2-based augmentation method providing realisticfake coins for rare classes. The dataset will be published in http://turcoins.Conference Object Citation - WoS: 9Citation - Scopus: 6Recognizing Non-Manual Signs in Turkish Sign Language(IEEE, 2019) Gökberk, Berk; Akarun, Lale; Aktaş, Müjde; 02.02. Department of Computer Engineering; 02. Faculty of Engineering; 01. MEF UniversityRecognition of non-manual components in sign language has been a neglected topic, partly due to the absence of annotated non-manual sign datasets. We have collected a dataset of videos with non-manual signs, displaying facial expressions and head movements and prepared frame-level annotations. In this paper, we present the Turkish Sign Language (TSL) non-manual signs dataset and provide a baseline system for non-manual sign recognition. A deep learning based recognition system is proposed, in which the pre-trained ResNet Convolutional Neural Network (CNN) is employed to recognize question, negation side to side and negation up-down, affirmation and pain movements and expressions. Our subject independent method achieves 78.49% overall frame-level accuracy on 483 TSL videos performed by six subjects, who are native TSL signers. Prediction results of consecutive frames are filtered for analyzing the qualitative results.Conference Object Citation - Scopus: 4Multi-View Reconstruction of 3d Human Pose With Procrustes Analysis(IEEE, 2019) Gökberk, Berk; Gökberk, Berk; Akarun, Lale; Temiz, Hüseyin; Gokherk, Berk; 02.02. Department of Computer Engineering; 02. Faculty of Engineering; 01. MEF UniversityRecovery of 3D human pose from cameras has been the subject of intensive research in the last decade. Algorithms that can estimate the 3D pose from a single image have been developed. At the same time, many camera environments have an array of cameras. In this paper, after aligning the poses obtained from single images using Procrustes Analysis, median filtering is utilized to eliminate outliers to find final reconstructed 3D body joint coordinates. Experiments performed on the CMU Panoptic, and Human3.6M databases demonstrate that the proposed system achieves accurate 3D body joint reconstructions. Additionally, we observe that camera selection is useful to decrease the system complexity while attaining the same level of reconstruction performance.