In the visual-spatial modality, signers indicate old, new, or contrastive information using certa... more In the visual-spatial modality, signers indicate old, new, or contrastive information using certain syntactic, prosodic, and morphological strategies. Even though information structure has been described extensively for many sign languages, the flow of information in the narrative discourse remains unexplored in Turkish Sign Language (TİD). This study aims to describe aboutness subject topic constructions in TİD narratives. We examined data from six adult native signers of TİD and found that TİD signers mainly used nominals for reintroduced aboutness subject topics. The optional and rare non-manual markers observed on reintroduced topics mainly included squint, brow raise, and backward head tilt. Maintained aboutness subject topics, which have higher referential accessibility, were often omitted and tracked with zero anaphora. Finally, we found that constructed action is more frequently present on the predicates of clauses with a maintained aboutness subject topic than with a reintr...
2021 IEEE International Conference on Image Processing (ICIP), 2021
In end-to-end optimized learned image compression, it is standard practice to use a convolutional... more In 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.
2021 IEEE International Conference on Image Processing (ICIP), 2021
It has become a standard practice to use the convolutional networks (ConvNet) with RELU non-linea... more It 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.
When comparing learned image/video restoration and compression methods, it is common to report pe... more When comparing learned image/video restoration and compression methods, it is common to report peak-signal to noise ratio (PSNR) results. However, there does not exist a generally agreed upon practice to compute PSNR for sets of images or video. Some authors report average of individual image/frame PSNR, which is equivalent to computing a single PSNR from the geometric mean of individual image/frame meansquare error (MSE). Others compute a single PSNR from the arithmetic mean of frame MSEs for each video. Furthermore, some compute the MSE/PSNR of Y-channel only, while others compute MSE/PSNR for RGB channels. This paper investigates different approaches to computing PSNR for sets of images, single video, and sets of video and the relation between them. We show the difference between computing the PSNR based on arithmetic vs. geometric mean of MSE depends on the distribution of MSE over the set of images or video, and that this distribution is task-dependent. In particular, these two methods yield larger differences in restoration problems, where the MSE is exponentially distributed and smaller differences in compression problems, where the MSE distribution is narrower. We hope this paper will motivate the community to clearly describe how they compute reported PSNR values to enable consistent comparison. Index Terms-PSNR, MSE, arithmetic mean, geometric mean, RGB-PSNR, Y-channel PSNR 1 O.K. and C.K. are supported by AI Fellowships provided by the Koç University-Turkish Is Bank (KUIS) AI Center. 2 A.M.T. acknowledges support from TUBITAK 1001 project 217E033, TUBITAK 2247-A National Leader Researchers Award (No. 120C156), and Turkish Academy of Sciences (TUBA). 3 Z.D. acknowledges support by the TUBITAK 2232 International Fellowship for Outstanding Researchers Award (No. 118C337).
Uploads
Papers by Onur Keleş