DR. ÖĞR. ÜYESİ HACER YALIM KELEŞ    
Name : HACER
Surname : YALIM KELEŞ
E-Mail : hkeles@ankara.edu.tr
Phone Number : +90-312-2033300/1756
Title : DR. ÖĞR. ÜYESİ
Unit : FACULTY OF ENGINEERING
Department : COMPUTER ENGINEERING DEPARTMENT
Personal Information

PhD Degree in Computer EngineeringMiddle East Technical University, (2010)

MS Degree in Computer Engineering: Middle East Technical University, (2005)

BSc in Computer Engineering: Middle East Technical University, (2002)


COURSES:

Undergraduate Level Courses:

COM101 Computer Programming I 

COM102 Computer Programming II

COM267 Data Structures

BLM438 Computer Graphics

COM448 Pattern Recognition

BLM492 Research Techniques I-II

Graduate Level Courses:

Deep Learning

Machine Learning

Computer Vision


RESEARCH INTERESTS

I have been coordinating the Computer Vision and Machine Learning (CVML) Lab since 2015. We primarily work with recent deep learning methods on various computer vision problems. Recently, we are working on Turkish Sign Language Recognition problem using video streams and deep learning models. In this context, we are gathering a large vocabulary TSL dataset with relevant annotations; this research is funded by TUBITAK. We are also currently working on a research using Generative Adversarial Networks (GANs), which is funded by Ankara University SRP. Moreover, we are making research on aerial scene recognition problem using deep learning models. In addition to these, I am currently interested in the speech coding problem.

In the last couple of years, I have been working with the students in their final years during their Undergraduate Graduation Projects with different computer vision problems using deep learning algorithms. Recently, one of these projects, which is a moving robot project that detects foreign objects on the ground, is supported by TUSAŞ-TAI. 


JOURNAL PUBLICATIONS    

Özge Mercanoğlu Sincan and Hacer Yalım Keleş"AUTSL: A Large Scale Multi-modal Turkish Sign Language Dataset and Baseline Methods", Preprint: arXiv:2008.00932. IEEE Access (in print), 2020, (SCI-Exp).

Özlem Şen and Hacer Yalım Keleş, "On the Evaluation of CNN Models in Remote Sense Scene Classification Domain", PFG-Journal of Photogrammetry, Remote Sensing and Geoinformation Science (in print), 2020, (SCI-Exp)

Anıl Osman Tur and Hacer Yalım Keleş, "Evaluation of Hidden Markov Models using Deep CNN Features in Isolated Sign Recognition", Preprint: arXiv: 2006.11183. Multimedia Tools and Applications (in review), 2020, (SCI-Exp).

Yahya Doğan and Hacer Yalım Keleş, "Semi-supervised Image Attribute Editing using Generative Adversarial Networks", Neurocomputing Journal, 2020(SCI) DOI: 10.1016/j.neucom.2020.03.071. (Preprint: arXiv:1907.01841).

Long Ang Lim and Hacer Yalım Keleş, "Learning Dense Contextual Features for Semantic Segmentation", Commun.Fac.Sci.Univ.Ank.Series A2-A3, vol. 62(1), pp. 26-34, 2020 (TR Index). 

Long Ang Lim and Hacer Yalım Keleş,  "Learning Multi-scale Features for Foreground Segmentation", Pattern Analysis and Applications (2019),  DOI: 10.1007/s10044-019-00845-9. (Preprint of this paper is available on ArXiv; arXiv:1808.01477.) (SCI-Expanded)

Hacer Yalım Keleş, Jan Rozhon, H. Gokhan Ilk, Miroslav Vosnak, "DeepVoCoder: A CNN model for compression and coding of narrow band speech", IEEE Access, June 2019, DOI: 10.1109/ACCESS.2019.2920663(SCI-Expanded)

Long Ang Lim and Hacer Yalım Keleş,  "Foreground Segmentation using Convolutional Neural Networks for Multiscale Feature Encoding", Pattern Recognition Letters 112, pp. 256-262, 2018 (SCI-Expanded)

Long Ang Lim and Hacer Yalım Keleş,  "Foreground Segmentation using a Triplet Convolutional Neural Network for Multiscale Feature Encoding"arXiv:1801.02225, 2018.

Hacer Yalım Keleş“Embedding Parts in Shape Grammars Using a Parallel Particle Swarm Optimization Method on GPUs”, AIEDAM, 32:3, pp. 256-268, 2018 (SCI-Expanded).

Özge Mercanoğlu Sincan, Hacer Yalım Keleş, Süleyman Tosun, "Moving Object Detection and Classification in Surveillance Systems Using Moving Cameras"Commun. Fac. Sci. Univ. Ank. Series A2-A3, 60(2), pp. 63-82, 2018 (TR Index)

Özge Mercanoğulu Sincan, Hacer Yalım KeleşYağmur Kır, Adnan Kusman, Bora Başkak “Person Identification Using Functional Near-Infrared Spectroscopy Signals Using a Fully Connected Deep Neural Network”, Commun. Fac. Sci. Univ. Ank. Series A2-A3, 59(2), pp. 57-65, 2017 (TR Index).

Nergis Pervan, Hacer Yalım Keleş“Sentiment Analysis Using a Random Forest Classifier on Turkish Web Comments”, Commun. Fac. Sci. Univ. Ank. Series A2-A3, 59(2), pp. 69-79, 2017 (TR Index).

Hacer Yalım Keleş“Acceleration of the Edge Strength Function on GPU Using CUDA”, Commun. Fac. Sci. Univ. Ank. Series A2-A3, 58(2), pp. 49-68, 2016 (TR Index)

Hacer Yalım Keleş, Sibel Tarı, “A Robust Method for Scale Independent Detection of Curvature-based Criticalities and In tersections in Line Drawings”, Pattern Recognition, 48(1), pp. 140-155, 2015 (SCI).

Hacer Yalım Keleş, Mine Özkar, Sibel Tarı, “Weighted Shapes for Embedding Perceived Wholes”Environment and Planning: B. , Volume 39, Issue 2, pp 360-375, 2012 (SSCI).

Hacer Yalım Keleş, Mine Özkar, Sibel Tarı, “Embedding Shapes without Predefined Parts”Environment and Planning: B, Volume 37, Issue 4, pp 664-681, 2010 (SSCI).

Hacer Yalım Keleş, Alphan Es, Veysi İşler, “Acceleration of Direct Volume Rendering with Programmable Graphics Hardware”, The Visual Computer: International Journal of Computer Graphics, Volume 23, Issue 1, pp 15-24, November, 2006 (SCI-Expanded).


CONFERENCES/SYMPOSIUMS:

Sinan Gençoğlu and Hacer Yalım Keleş, "Sign Language Video Synthesis using Skeleton Sequence", (Accepted) 28'th IEEE Conf. on Signal Processing and Communications Applications, 5-7 October 2020, Gaziantep, Turkey.

Özge Mercanoğlu Sincan, Sinan Gencoglu, Mert Bacak, Hacer Yalım Keleş, "Hand and Face Segmentation with Deep Convolutional Networks using Limited Labelled Data", 3rd International Symposium on Multidisciplinary Studies and Innovative Technologies (ISMSIT), IEEE, Ankara, Turkey, 11-13 October 2019.

Yahya Doğan, Hacer Yalım Keleş, "Stability and Diversity in Generative Adversarial Networks",  27'th IEEE Conference on Signal Processing and Applications, 24-26 April 2019, Sivas, Turkey.

Anıl Osman Tur, Hacer Yalım Keleş, "Isolated Sign Recognition with a Siamese Neural Network of RGB and Depth Streams", IEEE EUROCON 2019, 01-07 July 2019, Novi Sad, Serbia.

Özge Mercanoğlu Sincan, Anıl Osman Tur, Hacer Yalım Keleş, "Isolated Sign Language Recognition with Multi-scale Features using LSTM", 27'th IEEE Conference on Signal Processing and Applications, 24-26 April 2019, Sivas, Turkey.

Özlem Şen, Hacer Yalım Keleş, "Scene Recognition with Deep Learning Methods using Aerial Images", 27'th IEEE Conference on Signal Processing and Applications, 24-26 April 2019, Sivas, Turkey.

Nergis Pervan, Hacer Yalım Keleş, "A Dilated CNN Based Sentiment Analysis Using a Novel Corpus of Customer Product Reviews in Turkish Language", DTSS 2018, Ankara, October 24-26, 2018.

Elit Cenk Alp, Hacer Yalım Keleş, "A Comparative Study of HMMs and LSTMs on Action Classification with Limited Training Data", In the Proceedings of IntelliSys 2018, pp. 564-570, London, 2018.

Hacer Yalım Keleş, (Invited Speaker), “Utilization of Deep Learning Methods in Computer Graphics Domain”, Eurasiagraphics 2017, November 4-5, 2017. 

Hacer Yalım Keleş, (Keynote Speaker), “Recent Trends in Machine Learning and Computer Vision”, ICTACSE 2017, November 10-11, 2017.

Nergis Pervan, Hacer Yalım Keleş, “Sentiment Analysis Using a Random Forest Classifier on Turkish Web Comments”, In the Proceedings of ICTACSE 2017, Ankara, November 10-11, 2017.

Elit C. Alp, Hacer Yalım Keleş, "Action Recognition Using MHI Based Hu Moments with HMMs", Accepted in 17’th EUROCON 2017, 6-8 July2017, Ohrid, Macedonia.

Hacer Yalım Keleş, "Embedding Parts in Sketches Using a Parallel Evolutionary Approach", In 33’rd Education and Research in Computer Aided Architectural Design in Europe (eCAADe) Conference, 2015, Vienna, Austria.

Özge Mercanoglu Sincan,Vahid Babaei Ajabshir, Hacer Yalım Keleş, Suleyman Tosun, "Moving Object Detection by a Mounted Moving Camera", In EUROCON, 2015, Salamanca, Spain. 

Hacer Yalım Keleş, Erkin Tunca, Yücel Doğan, Fatih Keleş, Halit Oğuztüzün, “SOYA: Simülasyon Ortak Yaratım Aracı”, 6’th Defence Technologies Congress, June 2012, Ankara, Turkey.

Hacer Yalım Keleş, Mine Özkar, Sibel Tarı, “Revisiting Shape Embedding”, 27’th Education and Research in Computer Aided Architectural Design in Europe (eCAADe) Conference, 16-19, September 2009, İstanbul, Turkey.

Gökçe Yıldırım, Hacer Yalım Keleş, Veysi İşler, “Three Dimensional Smoke Simulation on Programmable Graphics Hardware”, International Conference on Parallel Computational Fluid Dynamics, 21-24, May 2007, Turkey.

Alphan Es, Hacer Yalım Keleş, Veysi İşler, “Accelerated Volume Rendering with Homogeneous Region Encoding using EACD on GPU”, 6’th Eurographics Symposium on Parallel Graphics and Visualization, 11-12, May 2006, Braga.


Supervised Master'sThesis (Completed):

  • Nergis PervanSemantic Inference from Turkish Texts Using Deep Learning Approaches, 2019.
  • Long Ang LimUtilization of Supervised Multi-scale Convolutional Neural Networks Approach for Foreground Segmentation, 2018.
  • Elit Cenk AlpAutomatic Activity Classification from Time-Sequenced Images Using Machine Learning Methods, 2018.
  • Özge Mercanoğlu SincanMoving Object Detection and Classification Using Moving Camera, 2015. (Co-advisor: Prof. Dr. Süleyman Tosun)
Ph.D. Thesis Students

  • Özge Mercanoğlu Sincan
  • Yahya Doğan
  • Özlem Şen
  • Zeynep Özdemir
  • Elit Cenk Alp
  • Nergis Pervan

PROJECTS:

  • TÜBİTAK ARDEB 3501 (June 2018, June 2020 ): Derin Öğrenme Yöntemleri Kullanılarak Geniş Dağarcıklı Türkçe İşaret Dili Tanıma Sisteminin Modellenmesi  (PI)
  • A.Ü. BAP - 18L0443010 (May 2018, May 2020 ): Üretken Çekişmeli Ağlarda Gizli Unsur Kodlayıcı ile Çıktı İmgesi Arasındaki İlişkinin Hesaplamalı Modellenmesi - (PI)
  • A.Ü. BAP - 15H0443009 (2015-2016): Eskizlerde Parça Gömme Probleminin CUDA Mimarisinde Optimizasyon Yoluyla Çözümü - (PI)
  • TÜBİTAK 1000 (2014-2017): Ankara Üniversitesi’nde Nitelikli Araştırma Projesi Üretim Altyapısının Geliştirilmesi - (Researcher)
  • TÜBİTAK 1507 (2011-2013): SOYA: Simulasyon Ortak Yaratım Aracı-(Researcher)
  • T.C. Sanayi ve Ticaret Bakanlığı (2010-2011): Ortak Test - (PI)
  • TÜBİTAK ARDEB 1001 (2008-2010): İki Boyutlu Desenlerde Görsel Düşünme Ve Tasarlama Süreçlerinin Şekil Cebiri Kullanarak Hesaplamalı Modellemesi-(Ph.D. Scholar)

AWARDS:

(PhD Thesis) Thesis of the Year Award, METU Prof. Dr. Mustafa Parlar Education and Research Foundation, 2010.