Publications with Fulltext
Permanent URI for this collectionhttps://hdl.handle.net/20.500.14288/6
Browse
9 results
Search Results
Publication Open Access A deep learning approach for data driven vocal tract area function estimation(Institute of Electrical and Electronics Engineers (IEEE), 2018) Department of Computer Engineering; Department of Electrical and Electronics Engineering; Erzin, Engin; Asadiabadi, Sasan; Faculty Member; Department of Computer Engineering; Department of Electrical and Electronics Engineering; College of Sciences; Graduate School of Sciences and Engineering; 34503; N/AIn this paper we present a data driven vocal tract area function (VTAF) estimation using Deep Neural Networks (DNN). We approach the VTAF estimation problem based on sequence to sequence learning neural networks, where regression over a sliding window is used to learn arbitrary non-linear one-to-many mapping from the input feature sequence to the target articulatory sequence. We propose two schemes for efficient estimation of the VTAF; (1) a direct estimation of the area function values and (2) an indirect estimation via predicting the vocal tract boundaries. We consider acoustic speech and phone sequence as two possible input modalities for the DNN estimators. Experimental evaluations are performed over a large data comprising acoustic and phonetic features with parallel articulatory information from the USC-TIMIT database. Our results show that the proposed direct and indirect schemes perform the VTAF estimation with mean absolute error (MAE) rates lower than 1.65 mm, where the direct estimation scheme is observed to perform better than the indirect scheme.Publication Open Access On training sketch recognizers for new domains(Institute of Electrical and Electronics Engineers (IEEE), 2021) Yeşilbek, Kemal Tuğrul; Department of Computer Engineering; Sezgin, Tevfik Metin; Faculty Member; Department of Computer Engineering; College of Engineering; 18632Sketch recognition algorithms are engineered and evaluated using publicly available datasets contributed by the sketch recognition community over the years. While existing datasets contain sketches of a limited set of generic objects, each new domain inevitably requires collecting new data for training domain specific recognizers. This gives rise to two fundamental concerns: First, will the data collection protocol yield ecologically valid data? Second, will the amount of collected data suffice to train sufficiently accurate classifiers? In this paper, we draw attention to these two concerns. We show that the ecological validity of the data collection protocol and the ability to accommodate small datasets are significant factors impacting recognizer accuracy in realistic scenarios. More specifically, using sketch-based gaming as a use case, we show that deep learning methods, as well as more traditional methods, suffer significantly from dataset shift. Furthermore, we demonstrate that in realistic scenarios where data is scarce and expensive, standard measures taken for adapting deep learners to small datasets fall short of comparing favorably with alternatives. Although transfer learning, and extensive data augmentation help deep learners, they still perform significantly worse compared to standard setups (e.g., SVMs and GBMs with standard feature representations). We pose learning from small datasets as a key problem for the deep sketch recognition field, one which has been ignored in the bulk of the existing literature.Publication Open Access Computational modeling of organisational learning by self-modeling networks(Elsevier, 2022) Treur, Jan; Roelofsma, Peter H. M. P.; Department of Computer Engineering; Canbaloğlu, Gülay; Department of Computer Engineering; Graduate School of Sciences and EngineeringWithin organisational learning literature, mental models are considered a vehicle for both individual learning and organizational learning. By learning individual mental models (and making them explicit), a basis for formation of shared mental models for the level of the organization is created, which after its formation can then be adopted by individuals. This provides mechanisms for organizational learning. These mechanisms have been used as a basis for an adaptive computational network model. The model is illustrated by a not too complex but realistic case study.Publication Open Access Stressed or just running? differentiation of mental stress and physical activity by using machine learning(TÜBİTAK, 2022) Department of History; Can, Yekta Said; Department of History; College of Social Sciences and HumanitiesRecently, modern people have excessive stress in their daily lives. With the advances in physiological sensors and wearable technology, people's physiological status can be tracked, and stress levels can be recognized for providing beneficial services. Smartwatches and smartbands constitute the majority of wearable devices. Although they have an excellent potential for physiological stress recognition, some crucial issues need to be addressed, such as the resemblance of physiological reaction to stress and physical activity, artifacts caused by movements and low data quality. This paper focused on examining and differentiating physiological responses to both stressors and physical activity. Physiological data are collected in the laboratory environment, which contain relaxed, stressful and physically active states and they are differentiated successfully by using machine learning.Publication Open Access Enhancing local linear models using functional connectivity for brain state decoding(IGI Global, 2013) Fırat, Orhan; Özay, Mete; Önal, Itır; Vural, Fatoş T. Yarman; Department of Psychology; Öztekin, İlke; PhD Student; Department of Psychology; College of Social Sciences and HumanitiesThe authors propose a statistical learning model for classifying cognitive processes based on distributed patterns of neural activation in the brain, acquired via functional magnetic resonance imaging (fMRI). In the proposed learning machine, local meshes are formed around each voxel. The distance between voxels in the mesh is determined by using functional neighborhood concept. In order to define functional neighborhood, the similarities between the time series recorded for voxels are measured and functional connectivity matrices are constructed. Then, the local mesh for each voxel is formed by including the functionally closest neighboring voxels in the mesh. The relationship between the voxels within a mesh is estimated by using a linear regression model. These relationship vectors, called Functional Connectivity aware Mesh Arc Descriptors (FC-MAD) are then used to train a statistical learning machine. The proposed method was tested on a recognition memory experiment, including data pertaining to encoding and retrieval of words belonging to ten different semantic categories. Two popular classifiers, namely k-Nearest Neighbor and Support Vector Machine, are trained in order to predict the semantic category of the item being retrieved, based on activation patterns during encoding. The classification performance of the Functional Mesh Learning model, which range in 62-68% is superior to the classical multi-voxel pattern analysis (MVPA) methods, which range in 40-48%, for ten semantic categories.Publication Open Access The effects of odor and body posture on perceived duration(Frontiers, 2014) Schreuder, Eliane; Hoeksma, Marco R.; Smeets, Monique A. M.; Department of Psychology; Semin, Gün Refik; Researcher; Department of Psychology; College of Social Sciences and Humanities; 58066This study reports an examination of the internal clock model, according to which subjective time duration is influenced by attention and arousal state. In a time production task, we examine the hypothesis that an arousing odor and an upright body posture affect perceived duration. The experimental task was performed while participants were exposed to an odor and either sitting upright (arousing condition) or lying down in a relaxing chair (relaxing condition). They were allocated to one of three experimental odor conditions: rosemary (arousing condition), peppermint (relaxing condition), and no odor (control condition). The predicted effects of the odors were not borne out by the results. Self-reported arousal (SRA) and pleasure (PL) states were measured before, during (after each body posture condition) and postexperimentally. Heart rate (HR) and skin conductance were measured before and during the experiment. As expected, odor had an effect on perceived duration. When participants were exposed to rosemary odor, they produced significantly shorter time intervals than in the no odor condition. This effect, however, could not be explained by increased arousal. There was no effect of body posture on perceived duration, even though body posture did induce arousal. The results do not support the proposed arousal mechanism of the internal clock model.Publication Open Access Generalized Polytopic Matrix Factorization(Institute of Electrical and Electronics Engineers (IEEE), 2021) Department of Electrical and Electronics Engineering; Erdoğan, Alper Tunga; Tatlı, Gökcan; Faculty Member; Department of Electrical and Electronics Engineering; College of Engineering; Graduate School of Sciences and Engineering; 41624; N/APolytopic Matrix Factorization (PMF) is introduced as a flexible data decomposition tool with potential applications in unsupervised learning. PMF assumes a generative model where observations are lossless linear mixtures of some samples drawn from a particular polytope. Assuming that these samples are sufficiently scattered inside the polytope, a determinant maximization based criterion is used to obtain latent polytopic factors from the corresponding observations. This article aims to characterize all eligible polytopic sets that are suitable for the PMF framework. In particular, we show that any polytope whose set of vertices have only permutation and/or sign invariances qualifies for PMF framework. Such a rich set of possibilities enables elastic modeling of independent/dependent latent factors with combination of features such as relatively sparse/antisparse subvectors, mixture of signed/nonnegative components with optionally prescribed domains.Publication Open Access Mukayese: Turkish NLP strikes back(Association for Computational Linguistics (ACL), 2022) Kurtuluş, Emirhan; Göktoğan, Arda; Department of Computer Engineering; Yüret, Deniz; Safaya, Ali; Faculty Member; Department of Computer Engineering; Koç Üniversitesi İş Bankası Yapay Zeka Uygulama ve Araştırma Merkezi (KUIS AI)/ Koç University İş Bank Artificial Intelligence Center (KUIS AI); College of Engineering; Graduate School of Sciences and Engineering; 179996; N/AHaving sufficient resources for language X lifts it from the under-resourced languages class, but not necessarily from the under-researched class. In this paper, we address the problem of the absence of organized benchmarks in the Turkish language. We demonstrate that languages such as Turkish are left behind the state-of-the-art in NLP applications. As a solution, we present MUKAYESE, a set of NLP benchmarks for the Turkish language that contains several NLP tasks. We work on one or more datasets for each benchmark and present two or more baselines. Moreover, we present four new bench-marking datasets in Turkish for language modeling, sentence segmentation, and spell checking.Publication Open Access Selection for function: from chemically synthesized prototypes to 3D-printed microdevices(Wiley, 2020) Bachmann, Felix; Giltinan, Joshua; Codutti, Agnese; Klumpp, Stefan; Faivre, Damien; Department of Mechanical Engineering; Sitti, Metin; Faculty Member; Department of Mechanical Engineering; School of Medicine; College of Engineering; 297104Magnetic microswimmers are promising devices for biomedical and environmental applications. Bacterium flagella-inspired magnetic microhelices with perpendicular magnetizations are currently considered standard for propulsion at low Reynolds numbers because of their well-understood dynamics and controllability. Deviations from this system have recently emerged: randomly shaped magnetic micropropellers with nonlinear swimming behaviors show promise in sensing, sorting, and directional control. The current progresses in 3D micro/nanoprinting allow the production of arbitrary 3D microstructures, increasing the accessible deterministic design space for complex micropropeller morphologies. Taking advantage of this, a shape is systematically reproduced that was formerly identified while screening randomly shaped propellers. Its nonlinear behavior, which is called frequency-induced reversal of swimming direction (FIRSD), allows a propeller to swim in opposing directions by only changing the applied rotating field's frequency. However, the identically shaped swimmers do not only display the abovementioned swimming property but also exhibit a variety of swimming behaviors that are shown to arise from differences in their magnetic moment orientations. This underlines not only the role of shape in microswimmer behavior but also the importance of determining magnetic properties of future micropropellers that act as intelligent devices, as single-shape templates with different magnetic moments can be utilized for different operation modes.