Research Outputs
Permanent URI for this communityhttps://hdl.handle.net/20.500.14288/2
Browse
6 results
Search Results
Publication Restricted Cognitively-inspired deep learning approaches for grounded language learning(Koç University, 2021) Can, Ozan Arkan; Yüret, Deniz; 0000-0002-7039-0046; Koç University Graduate School of Sciences and Engineering; Computer Science and Engineering; 179996Publication Open Access Computerized counting of individuals in Ottoman population registers with deep learning(Springer, 2020) Department of History; Can, Yekta Said; Kabadayı, Mustafa Erdem; Faculty Member; Department of History; College of Social Sciences and Humanities; N/A; 33267The digitalization of historical documents continues to gain pace for further processing and extract meanings from these documents. Page segmentation and layout analysis are crucial for historical document analysis systems. Errors in these steps will create difficulties in the information retrieval processes. Degradation of documents, digitization errors and varying layout styles complicate the segmentation of historical documents. The properties of Arabic scripts such as connected letters, ligatures, diacritics and different writing styles make it even more challenging to process Arabic historical documents. In this study, we developed an automatic system for counting registered individuals and assigning them to populated places by using a CNN-based architecture. To evaluate the performance of our system, we created a labeled dataset of registers obtained from the first wave of population registers of the Ottoman Empire held between the 1840s–1860s. We achieved promising results for classifying different types of objects and counting the individuals and assigning them to populated places.Publication Open Access Self-organized residual blocks for image super-resolution(Institute of Electrical and Electronics Engineers (IEEE), 2021) Malik, J.; Kıranyaz, S.; Department of Electrical and Electronics Engineering; Tekalp, Ahmet Murat; Keleş, Onur; Faculty Member; Department of Electrical and Electronics Engineering; College of Engineering; Graduate School of Sciences and Engineering; 26207; N/AIt has become a standard practice to use the convolutional networks (ConvNet) with RELU non-linearity in image restoration and super-resolution (SR). Although the universal approximation theorem states that a multi-layer neural network can approximate any non-linear function with the desired precision, it does not reveal the best network architecture to do so. Recently, operational neural networks (ONNs) that choose the best non-linearity from a set of alternatives, and their “self-organized” variants (Self-ONN) that approximate any non-linearity via Taylor series have been proposed to address the well-known limitations and drawbacks of conventional ConvNets such as network homogeneity using only the McCulloch-Pitts neuron model. In this paper, we propose the concept of self-organized operational residual (SOR) blocks, and present hybrid network architectures combining regular residual and SOR blocks to strike a balance between the benefits of stronger non-linearity and the overall number of parameters. The experimental results demonstrate that the proposed architectures yield performance improvements in both PSNR and perceptual metrics.Publication Open Access Self-organized variational autoencoders (self-vae) for learned image compression(Institute of Electrical and Electronics Engineers (IEEE), 2021) Malik, J.; Kıranyaz S.; Department of Electrical and Electronics Engineering; Tekalp, Ahmet Murat; Keleş, Onur; Yılmaz, Mustafa Akın; Güven, Hilal; Faculty Member; Department of Electrical and Electronics Engineering; College of Engineering; Graduate School of Sciences and Engineering; 26207; N/A; N/A; N/AIn end-to-end optimized learned image compression, it is standard practice to use a convolutional variational autoencoder with generalized divisive normalization (GDN) to transform images into a latent space. Recently, Operational Neural Networks (ONNs) that learn the best non-linearity from a set of alternatives, and their “self-organized” variants, Self-ONNs, that approximate any non-linearity via Taylor series have been proposed to address the limitations of convolutional layers and a fixed nonlinear activation. In this paper, we propose to replace the convolutional and GDN layers in the variational autoencoder with self-organized operational layers, and propose a novel self-organized variational autoencoder (Self-VAE) architecture that benefits from stronger non-linearity. The experimental results demonstrate that the proposed Self-VAE yields improvements in both rate-distortion performance and perceptual image quality.Publication Restricted Training memory-constrained deep learning models using automatic dataflow-graph partitioning(Koç University, 2020) Qararyah, Fareed Mohammad Fareed; Erten, Didem Unat; 0000-0002-2351-0770; Koç University Graduate School of Sciences and Engineering; Computer Engineering; 219274Publication Restricted Transition based dependency parsing with deep learning(Koç University, 2018) Kırnap, Ömer; Yüret, Deniz; 0000-0002-7039-0046; Koç University Graduate School of Sciences and Engineering; Computer Science and Engineering; 179996