WoS İndeksli Yayınlar Koleksiyonu / WoS Indexed Publications Collection

Permanent URI for this collectionhttps://hdl.handle.net/20.500.11779/256

Browse

Browsing WoS İndeksli Yayınlar Koleksiyonu / WoS Indexed Publications Collection by Publication Category "Konferans Öğesi - Ulusal - Kurum Öğretim Elemanı"

Filter results by typing the first few letters
Now showing 1 - 4 of 4
  • Thumbnail Image
    Conference Object
    Citation - WoS: 3
    Citation - Scopus: 3
    Data Repair in Bs-Assisted Distributed Data Caching
    (IEEE, 2020) Kaya, Erdi; Haytaoğlu, Elif; Arslan, Şuayb Şefik
    In this paper, centralized and independent repair approaches based on device-to-device communication for the repair of the lost nodes have been investigated in a cellular network where distributed caching is applied whose fault tolerance is provided by erasure codes. The caching mechanisms based on Reed-Solomon codes and minimum bandwidth regenerating codes are adopted. The proposed approaches are analyzed in a simulation environment in terms of base station utilization load during the repair process. Based on the intuitive assumption that the base station is usually more costly than device-to-device communication, the centralized repair approach demonstrates a better performance than the independent repair approaches on the number of symbols retrieved from the base station. On the other hand, the centralized approach has not achieved a dramatic reduction in the number of symbols downloaded from the other devices.
  • Thumbnail Image
    Conference Object
    Citation - WoS: 1
    Citation - Scopus: 1
    Face Recognition With Local Zernike Moments Features Around Landmarks
    (IEEE, 2016) Gökmen, Muhittin; Basaran, Emrah
    In this paper, a new method that extracts the features from the complex Local Zernike Moments (LZM) images around facial landmarks is proposed. In this method, multiple grids which are in different sizes are located on landmarks and Phase-Magnitude (PM) histograms are calculated in each cells of these grids. The PM histograms are calculated for every component of LZM and the feature vectors are created by concatenating these histograms. By reducing the dimensionality of these vectors using Whitened Principle Component Analysis, more robust descriptors are constructed. It is shown that the state-of-the-art results are obtained in the experiments performed on FERET database using the proposed method. © 2016 IEEE.
  • Thumbnail Image
    Conference Object
    Impact of Hardware Sources on Feature Selection for Online Signature Verification
    (IEEE, 2020) Ayhan, Tuba; Orak, Remzi
    This work analyzes time series features gathered from a touchpad which is a part of online signature verification system. A DTW processing unit is implemented on FPGA to be used in time series analysis. To support different feature groups, this unit can be reconfigured without altering the memory structure. By using this reconfigurable unit, features are evaluated according to the area cost that they introduce. Moreover, a method to predict the value of features for classification is introduced. This way, minimum requirements to implement an online signature verification system on FPGA are partially obtained.
  • Thumbnail Image
    Conference Object
    Citation - WoS: 2
    Citation - Scopus: 2
    Turkish Broadcast News Transcription Revisited
    (IEEE, 2018) Saraçlar, Murat; Arısoy, Ebru
    Bu çalışmada yaklaşık on yıl önce gerçeklenen Türkçe haber programları için otomatik konuşma tanımayla yazılandırma sistemi güncel yöntemlerle yenilenerek aynı veri üzerindeki başarımı ölçülmüştür. Son yıllarda yapay sinir ağları temelli derin öğrenme yöntemleri konu¸sma tanıma hata oranlarında belirgin bir iyileşme sağlamıştır ve günümüzde yaygın olarak kullanılmaktadır. Bu bildiride geliştirilen konu¸sma tanıma sisteminin temel bileşenleri olan akustik ve dil modelleri için sinir ağları kullanılmıştır. Akustik modelleme için derin sinir a^gları hem çapraz entropi hem de ayırıcı dizi amaç işlevleriyle eniyilenmiştir. Ayrıca uzun süreli bağımlılıkları modellemek için yinelemeli sinir ağlarına benzer bir başarım gösteren ama daha çabuk eğitilebilen zaman gecikmeli sinir ağları kullanılmıştır. Daha sonra bunların ayırıcı eğitimle eniyilenmesi sonucunda en düşük hata oranlarına ulaşılmoştır. Dil modeli için ise yinelemeli sinir ağları kullanılmıştır. Bu yeni sinir ağları kullanan modeller ile kelime hata oranlarının yarılandığıve %10’un altına düştüğü gözlemlenmiştir.