Elektrik- Elektronik Mühendisliği

Permanent URI for this community

https://digitalarchive.library.bogazici.edu.tr/handle/123456789/11605

Browse

Browsing Elektrik- Elektronik Mühendisliği by Author "Akgül, Ceyhun Burak."

Now showing 1 - 4 of 4
  • Thumbnail Image
    Item
    Analysis of functional near infrared spectroscopy signals
    (Thesis (M.S.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2004., 2004.) Akgül, Ceyhun Burak.; Sankur, Bülent.
    In recent years, positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) have facilitated the monitoring of the human brain non-invasively, during functional activity. Nevertheless, the use of these systems remain limited since they are expensive, they cannot provide sufficient temporal detail and they are not very comfortable for the patient or the volunteer whose brain is monitored. Functional near infrared spectroscopy (fNIRS), on the other hand, is an emerging non-invasive modality which may be a remedy for the failures of the existing technologies. However, properly designed data analysis schemes for fNIRS have been missing. In this M.S. thesis, we intend to introduce a collection of signal processing methods in order to treat fNIRS data acquired during functional activity of the human brain. Along extensive hypothesis tests that characterized the statistical properties of the empirical data, we have described the signals in the time-frequency plane and partitioned the signal spectrum into several dissimilar subbands using an hierarchical clustering procedure. The proposed subband partitioning scheme is original and can easily be applied to signals other than fNIRS. In addition to these, we have adapted two different exploratory data analysis tools, namely, independent component analysis (ICA) and waveform clustering, to fNIRS short-time signals in order to learn generic cognitive activity-related waveforms, which are the counterparts of the brain hemodynamic response in fMRI. The periodicity analysis of the signals in the 30-250 mHz range validates that fNIRS measures indeed functional cognitive activity. Furthermore, as extensive ICA and waveform clustering experiments put into evidence, cognitive activity measured by fNIRS, reveals itself in a way very similar to the one measured by fMRI. These findings indicate that, in the near future, fNIRS shall play a more important role in explaining cognitive activity of the human brain.
  • Thumbnail Image
    Item
    Density-based shape descriptors and similarity learning for 3D object retrieval
    (Thesis (Ph.D.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2008., 2008.) Akgül, Ceyhun Burak.; Sankur, Bülent.; Schmitt, Francis.; Yemez, Yücel.
    Next generation search engines will enable query formulations, other than text, relying on visual information encoded in terms of images and shapes. Content-based retrieval research aims at developing search engines that would allow users to perform a query by similarity of content. This thesis deals with two fundamentals problems in content-based 3D object retrieval: (1) How to describe a 3D shape to obtain a reliable representative for the subsequent task of similarity search? (2) How to supervise the search process to learn inter-shape similarities for more effective and semantic retrieval? Concerning the first problem, we develop a novel 3D shape description scheme based on probability density of multivariate local surface features. We constructively obtain local characterizations of 3D points on a 3D surface and then summarize the resulting local shape information into a global shape descriptor. This conversion mechanism circumvents the correspondence problem between two shapes and proves to be robust and effective. Experiments that we have conducted on several 3D object databases show that density-based descriptors are very fast to compute and very effective for 3D similarity search. Concerning the second problem, we propose a similarity learning scheme that incorporates a certain amount of supervision into the querying process. Our approach relies on combining multiple similarity scores by optimizing a convex regularized version of the empirical ranking risk criterion. This score fusion approach to similarity learning is applicable to a variety of search engine problems using arbitrary data modalities. In this work, we demonstrate its effectiveness in 3D object retrieval.
  • Thumbnail Image
    Item
    Improving image captioning with language modeling regularizations
    (Thesis (M.S.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 2019., 2019.) Ulusoy, Okan.; Anarım, Emin.; Akgül, Ceyhun Burak.
    Inspired by the recent work in language modeling, we investigate the effects of a set of regularization techniques on the performance of a recurrent neural network based image captioning model. Using these techniques, we achieve 13 Bleu-4 points improvements over using no regularizations. We show that our model does not suffer from loss-evaluation mismatch and also connect the model performance to dataset properties by running experiments on MSCOCO dataset. Further, we propose two different applications for our image captioning model, namely human in the loop system and zero shot object detection. The former application further improves CIDEr score of our best model by 30 points using only the first two tokens of a reference sentence of an image. In the latter one, we train our image captioning model as an object detector which classifies each objects in an image without finding their location. The main advantage of this detector is that it does not require object locations during the training phase.
  • Thumbnail Image
    Item
    People detection in cluttered scenes
    (Thesis (M.S.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2010., 2010.) Öztürk, Serdar.; Sankur, Bülent.; Akgül, Ceyhun Burak.
    In this thesis, we have performed people detection in cluttered scenes. The people search operation in an image is performed by sliding a detection window and converting the content of each window to a feature vector. Dense feature representation of the detection window is obtained by dividing it into overlapping blocks and extracting local features of the blocks. These block features are concatenated to form the combined feature vector of the detection window. Feature vectors are obtained from windows with people and not containing people (negative samples), and used to train a linear SVM classifier. We have studied various types of features to use for people detection. First, we have performed people detection using Histogram of Oriented Gradients (HOG) using various combinations of values of HOG feature extraction parameters like block sizes, gradient operators, HOG bin numbers and normalization methods. In addition to HOG features, we have also studied other features like Gabor energies, block orientation vectors, skin color, projection profiles and cluster distances. In order to increase the performance and the reliability of the detection algorithm, various fusion techniques are applied at data, feature and decision levels. For example, HOG based detector scores are fused with Gabor based detector scores and improved detection scores are obtained. Also, same type detectors are varied by changing detector parameters and the detection scores of these detectors are combined. The performance of the algorithms is measured using different parameters and configurations, and results are compared using Detection Error Tradeoff plots.