Kadir A_20160804_res_tea
1 like471 views
Kadir A. Peker is an assistant professor in the Department of Computer Engineering at Melikşah University in Turkey. His research interests include machine learning, computer vision, and deep learning, with a focus on convolutional neural networks for image matching and comparison. He has over 17 years of experience in academia and industry, teaching courses in programming, machine learning, and computer vision.
Kadir A_20160804_res_tea
- 1. Kadir A. Peker 1 Kadir A. Peker, PhD. Tel: +90-536-298-4324 e-mail: kadir.peker@gmail.com Assistant Professor Department of Computer Engineering Melikşah University Talas, Kayseri, TURKEY Google Scholar (Kadir Aşkın Peker): h-index: 17. i-10 index: 25. Citations: 822. LinkedIn: http://tr.linkedin.com/in/kadirapeker ResearchGate RG Score: 10.48 Quara: Kadir A. Peker Research Interests Current Interests: Machine Learning. Computer Vision. Deep Learning. Deep Learning: o Deep networks, especially Convolutional Neural Networks: Multi-image input CNN architectures for image matching, comparison and multi-modal recognition: Match image pairs (Are these images of the same place?) Compare image pairs (Which is a prettier/older/more trustworthy/etc. face?) Merge image pairs (IR + visible light object recognition). o Input image optimization for a deeper understanding of CNNs and for artistic image generation (deep dreams). o Other: Recurrent neural networks. Theoretical understanding and applications to new domains. Unsupervised or semi-supervised learning. Image comparison/matching, scene and object recognition using SIFT etc. descriptors. Text mining and NLP using deep learning techniques. Prior Research Areas: Video and image content analysis, retrieval, browsing and summarization. Mobile applications, location based services, web-based applications. Visualization, computer art. Teaching Courses Taught: Introduction to Programming (Java, Python), Object Oriented Programming (Java), Machine Learning (both graduate and undergraduate), Computer Vision, Linear Algebra and Applications (including Matlab) , Signals and Systems, Web Programming (Javascript + PHP), Discrete Mathematics, Computer Organization. Consistently high student ratings and very positive student feedback. Most popular senior project advisor. Similar popularity in offered elective courses. Implemented some novel teaching ideas to handle failing students, got positive results.
- 2. Kadir A. Peker 2 Education 1996 – 2001 New Jersey Institute of Technology, Newark NJ, Ph.D. in Electrical and Computer Engineering Research Assistant in New Jersey Center for Multimedia Research (NJCMR) Graduation date: January 2001 Thesis Title: Video Indexing and Summarization Using Motion Activity 1994 – 1996 Rutgers University, New Brunswick NJ, MS in Electrical Engineering – (Signal Processing and Communications Area) 1989 – 1993 Bilkent University, Ankara Turkey, BS in Electrical Engineering Honors Ranked 29th among over 500,000 people in the nationwide university entrance exam in Turkey in 1989 All education was on scholarships earned from the government or private organizations Professional Experience Assistant Professor, (2009 – 2016) Department Chair (2009 – 2011) Department of Computer Engineering, Melikşah University, Kayseri, Turkiye Founding chair of the department: Designed the curriculum, department goals and priorities; for both the undergraduate and the post graduate programs. Extensive research of bylaws, educational issues, etc. regarding higher education. Member of the university senate and several key committees. Actively participated in developing the organizational and academic structures of the university. Established a Computer Vision and Machine Learning Lab: Most preferred thesis advisor in the department: Established Computer Vision and Machine Learning research area as arguably the most sought after field in the department by offering: Computer Vision and Machine Learning graduate courses; Introductory undergraduate machine learning courses; Innovative senior projects using OpenCV, regression/classification, etc. Some funded projects: Video browsing using face recognition (PI, with Arçelik A.Ş. Government supported). Medical image analysis for stereotactic brain surgery (consultant, government support). Trojan detection in VLSI chips using image analysis (Researcher, defense project).
- 3. Kadir A. Peker 3 Recent Theses and Project Topic Samples: Recognizing face attributes such as attractiveness, trustworthiness using a pair-wise comparison CNN. IR and visible light object recognition using a merged CNN architecture and comparison to using separate CNNs. A classification approach to SIFT matching for day and night location matching. Generating artificial images that maximize classification score of a CNN but that are usually meaningless to human eyes, for a deeper understanding of CNNs. Fabric defect detection using deep neural networks, SIFT descriptors. Mining literary texts, twitter for stylistic analysis. Broadcast video browsing based on person recognition in video images. Teaching: Consistently high student evaluations: Feedbacks such as “all the courses should be taught by him”, “we would like to become like you, what should we study”, “I want to change to Turkish section to be in your class”, etc. Some topics taught: Introduction to Programming (Java, Python), Object Oriented Programming (Java), Machine Learning (both graduate and undergraduate), Computer Vision, Linear Algebra and Applications (including Matlab), Web Programming (Javascript), Signals and Systems, Computer Organization. Assistant Professor, (2006 – 2009) Computer Technology and Information Systems Department, Bilkent University, Ankara, Turkiye Taught discrete mathematics, web technologies, signals and systems, introduction to electrical engineering. Research on fault detection from sensor data; image matching in street pictures using SIFT. Departmental duties on accreditation, curriculum. Member Technical Staff, Cambridge, MA (2003 – 2006) Murray Hill, NJ (2001 – 2003) Mitsubishi Electric Research Laboratories (MERL), USA Data mining for anomaly detection, equipment condition monitoring, and prognostics Studied anomaly detection problem; used locally weighted regression methods, spectral clustering, and other statistical techniques for fault detection and diagnosis in HVAC (heating, ventilation, air conditioning) equipment.
- 4. Kadir A. Peker 4 Video Analysis, Browsing, and Summarization Conducted research on consumer video analysis for browsing and summarization: Developed techniques for browsing and summarization of video using face detection. Applied tracking and clustering techniques for temporal segmentation of video. Implemented a face detection plug-in in C++ for an open source video editing software (Virtual Dub). Developed smart fast-forwarding based video browsing techniques. Developed novel visual complexity measures and extraction techniques for mpeg-1 and mpeg-2 encoded video. Implemented demos using Java, JMF, Visual C++. Contributed to the transfer of technology to real consumer products such as DVD-Harddisk recorder, and TV with hard disk storage. Leading role in project planning. Worked on usability issues; user interface development for PVR’s incorporating smart video browsing and summarization. Participated in conducting and evaluation of a usability study. Studied evaluation of video summaries in an application context. Worked on audio-visual event mining techniques using time series analysis methods. Supported audio analysis and classification research and development for sports video summarization on consumer electronics platforms. Home Networking Participated in standards activities in home networking area, including UPnP and OSGi. Implemented a UPnP Media Server on a PC platform. Developed home media services for an advanced digital network project. Internship at Mitsubishi Electric Research Labs, Murray Hill, NJ (1999 – 2000) Research: Developed motion activity based methods of video summarization, emphasizing feasibility of algorithms and implementation on consumer electronics platforms. Participated in and supported Mpeg-7 standards activity at the lab. Conducted a psychophysical experiment on perception and automatic measurement of motion activity in video. Development: Developed and maintained demo software for retrieval, browsing, and summarization of video using motion activity and color, using Java, JDBC and MS Access. Developed a web-based Applet-Servlet version of the video-browsing tool for the Mitsubishi press demo. Research Assistantship at NJIT/NJCMR (1997 – 1998) Worked on video motion analysis using optical flow, multiresolution representation of motion information, similarity measures, and clustering using motion characteristics. Implementation using C and Matlab.
- 5. Kadir A. Peker 5 Participated in Mpeg-7 standards activities in the Motion Activity Description area. Rutgers Vision Lab (1995) Implemented a model of human visual system motion detection model on Matlab. Bilkent University (1993) Generated fractal clouds for computer graphics using 1/f filtering of white noise in Matlab. Added new features to the Soundtool in XWindow (Unix) using XView C libraries. Programming Skills ● Programming Languages: Java, Python, Matlab, C/C++, Pascal, Assembly. ● Have actively been developing code for research, projects and personal interests/hobbies, such as: ● Image classification, image optimization using convolutional neural networks (Matlab). ● Video content analysis and face recognition, variable speed video playback (OpenCV, python, C#-emguCV, Matlab). ● Text mining – n-gram statistics, classification, etc. (Java, python). ● Image matching using SIFT (python, PHP+javascript-web app). ● Solving peg-solitaire using DFS and BFS under memory and time constraints (python, Java, C++). ● Implementing a stereotactic measurement plug-in for the Slicer 3D medical imaging software (Python, VTK/ITK, some C++). ● A large number of course projects and labs developed for introduction to programming, object oriented programming, web programming, machine learning, computer vision courses and senior projects. ● Participating on coding competition sites such as codewars. ● Self-taught programmer since high school (1986-89) when computers and learning resources were almost non-existent. First hobby projects on projectile simulation, animation and simple games. Professional Activities Reviewer and panelist for Tübitak and Teydeb projects (The Scientific and Technological research council of Turkey). Panelist for KOSGEB (Small and Medium Industry Development Organization of Turkey). Reviewer for IEEE CSVT, IEEE Trans MM, KAIS (Springer), JEI, several conferences.
- 6. Kadir A. Peker 6 Technical program committee member, ICME03, SPIE EI SPIE EI - Multimedia Content Access: Algorithms and Systems (2005 and 2006). Co-organized special session on video summary evaluation, at Electronic Imaging 2006. Other Interests Visual arts, including computer graphics and photography; Music (play the Ney – a traditional Turkish wind instrument); Table tennis (racket with custom blade and rubbers) and other sports; Computer programming and mathematical puzzles. Beginner-intermediate level Japanese. Beginner level Arabic. Selected Publications 1. Nazlı Tekin, Kadir A. Peker, “Matching day and night location images using sift and logistic regression,” IEEE 23nd Signal Processing and Communications Applications Conference (SIU 2015), Malatya, Turkey, May 2015. 2. Kadir A. Peker, Gökhan Özsarı, “Contaminant and foreign fiber detection in cotton using Gaussian mixture model,” IEEE 8th Intr. Conf. on Application of Information and Communication Technologies (AICT 2014), Astana, Kazakhstan, Oct. 2014. 3. Fahriye Gemci, Kadir A. Peker, “Extracting Turkish tweet topics using LDA,” IEEE 8th International Conference on Electrical and Electronics Engineering (ELECO 2013), Bursa, Turkey, Nov. 2013. 4. Kadir A. Peker, “Binary SIFT: Fast Image Retrieval Using Binary Quantized SIFT Features,” IEEE Intl. Workshop on Content Based Multimedia Indexing (CBMI 2011), Madrid, Spain, June 2011. 5. R. Radhakrishnan, D. Nikovski, K. Peker, A. Divakaran, "A comparison between polynomial and locally weighted regression for fault detection and diagnosis of HVAC equipment," IEEE Industrial Electronic Conf. (IECON06). 6. Peker, K.A.; Otsuka, I.; Divakaran, A., “Broadcast video program summarization using face tracks,” IEEE Intl. Conference on Multimedia and Expo (ICME 2006), Toronto, Canada, Jul. 2006. 7. Forlines, C.; Peker, K.A.; Divakaran, A., “Subjective assessment of consumer video summarization (Invited Paper)”, SPIE Multimedia Content Analysis, Management, and Retrieval 2006, San Jose, CA. Jan 2006.
- 7. Kadir A. Peker 7 8. A. Divakaran, R. Radhakrishnan, K. A. Peker, “Blind summarization: content adaptive video summarization using time-series analysis,” SPIE Multimedia Content Analysis, Management, and Retrieval 2006, San Jose, CA. Jan 2006. 9. Peker, K.A.; Bashir, F., “Content-based video summarization using spectral clustering,” Intl. VLVB Workshop, Sep 2005. 10. Peker, K.A; Divakaran, A.; Lanning, T. “Browsing news and talk video on a consumer electronics platform using face detection,” SPIE Multimedia Systems and Applications VIII, Oct 2005. 11. Peker, K.A, “Subsequence Time Series (STS) Clustering Techniques for Meaningful Pattern Discovery,” IEEE Intl. Conf. on Integration of Knowledge Intensive Multi-Agent Systems (KIMAS 2005), Waltham, MA, April 2005. 12. Peker, K.A.; Divakaran, A., “Adaptive Fast Playback-Based Video Skimming Using a Compressed-Domain Visual Complexity Measure,” IEEE Int’l. Conf. on Multimedia and Expo (ICME), Taipei, June 2004. 13. Divakaran, A.; Peker, K.A.; Chang, S-F; Radhakrishnan, R.; Xie, L., "Video Mining: Pattern Discovery versus Pattern Recognition (Invited)", IEEE International Conference on Image Processing (ICIP), ISSN: 1522-4880, Vol. 4, pp. 2379-2382, October 2004 (Also MERL TR2004-127) 14. Divakaran, A.; Miyaraha, K.; Peker, K.A.; Radhakrishnan, R.; Xion, Z., "Video Mining using Combinations of Unsupervised and Supervised Learning Techniques", SPIE Conference on Storage and Retrieval for Multimedia Databases, Vol. 5307, pp. 235-243, January 2004 (Also MERL TR2004-007) 15. Peker, K.A.; Divakaran, A., "Framework for Measurement of the Intensity of Motion Activity of Video Segments", Journal of Visual Communications and Image Representation, Vol. 14, Issue 4, December 2003 (Also MERL TR2003-64). 16. Divakaran, A.; Peker, K.A.; Radharkishnan, R.; Xiong, Z.; Cabasson, R., "Video Summarization Using MPEG-7 Motion Activity and Audio Descriptors", in Video Mining, Eds. Rosenfeld, A.; Doermann, D.; DeMenthon, D., October 2003, Kluwer Academic Publishers. (Also MERL TR2003-34) 17. Divakaran, A.; Radhakrishnan, R.; Peker, K.A., "Motion Activity-Based Extraction of Key-Frames from Video Shots", IEEE International Conference on Image Processing (ICIP), ISSN: 1522-4880, Vol. 1, pp. 932-935, September 2002. 18. Sahinoglu, Z.; Peker, K.A.; Matsubara, F.; Cukier, J., "A Mobile Network Architecture with Personalized Instant Information Access", IEEE International Conference on Consumer Electronics (ICCE), pp. 34-35, June 2002. 19. Bhatti, G.M.; Sahinoglu, Z.; Peker, K.A.; Guo, J.; Matsubara, F., "A TV-Centric Home Network to Provide a Unified Access to UPnP and PLC Domains", IEEE International Workshop on Networked Appliances (IWNA), pp. 234-242, January 2002.
- 8. Kadir A. Peker 8 20. Peker, K.A.; Cabasson, R.; Divakaran, A., "Rapid Generation of Sports Video Highlights Using the MPEG-7 Motion Activity Descriptor", SPIE Conference on Storage and Retrieval for Media Databases, Vol 4676, pps 318-323, January 2002. 21. Divakaran, A.; Radharkishnan, R.; Peker, K.A., "Video Summarization Using Descriptors of Motion Activity: A Motion Activity Based Approach to Key-Frame Extraction from Video Shots", Journal of Electronic Imaging, Vol. 10, Issue 4, pp. 909-916, October 2001. 22. Peker, Kadir A.; Divakaran, Ajay; Sun, Huifang, “Constant Pace Skimming and Temporal Sub-sampling of Video Using Motion Activity,” Proc. of IEEE International Conference on Image Processing (ICIP) 2001. 23. Kadir A. Peker, Ajay Divakaran, Thomas V. Papathomas, “Automatic Measurement of Intensity of Motion Activity of Video Segments,” Proc. SPIE 4315, Storage and Retrieval for Media Databases 2001, Jan. 2001. 24. Kadir A. Peker, A. Aydin Alatan and Ali N. Akansu, “Low-level Motion Activity Features for Semantic Characterization of Video,” Proc. of IEEE International Conference on Multimedia and Expo (ICME) 2000. Selected Patents 1. Descriptor for Spatial Distribution of Motion Activity in compressed video Ajay Divakaran, Kadir A. Peker, Huifang Sun. USP 6,600,784. Filed: February 2, 2000. Granted: July 29, 2003. 2. Method for summarizing a video using motion and color descriptors Ajay Divakaran, Kadir A. Peker, Huifang Sun. USP 6,697,523. Filed: August 9, 2000. Granted: February 24, 2004. 3. Summarizing videos using motion activity descriptors correlated with audio features Romain Cabasson, Kadir A. Peker, Ajay Divakaran USP 6,956,904. Filed: January 15, 2002. Granted: October 18, 2005. 4. Adaptively processing a video based on content characteristics of frames in the video Kadir A. Peker, Ajay Divakaran, Huifang Sun USP 7,003,154. Filed: November 17, 2000. Granted: February 21, 2006. 5. Pattern discovery in multi-dimensional time series using multi-resolution matching Kadir A. Peker USP 7,103,222. Filed: November 1, 2002. Granted: September 5, 2006. 6. Method for summarizing a video using motion descriptors Ajay Divakaran, Regunathan Radhakrishnan, Kadir A. Peker
- 9. Kadir A. Peker 9 USP 7,110,458. Filed: Apr 27, 2001. Granted: September 19, 2006. 7. Blind summarization of video content Ajay Divakaran, Kadir A. Peker USP 7,143,352. Filed: Nov 1, 2002. Granted: November 28, 2006. 8. Method for detecting short term unusual events in videos Ajay Divakaran, Ziyou Xiong, Regunathan Radhakrishnan, Kadir A. Peker, Koji Miyahara USP 7,327,885. Filed: June 30, 2003. Granted: February 5, 2008. 9. Video mining using unsupervised clustering of video content Ajay Divakaran, Kadir A. Peker USP 7,375,731. Filed: Nov 1, 2002. Granted: May 20, 2008. 10. Visual complexity measure for playing videos adaptively Kadir A. Peker, Ajay Divakaran USP 7,406,123. Filed: July 10, 2003. Granted: July 29, 2008. 11. Detecting and Diagnosing Faults in HVAC Equipment Daniel Nikovski, Ajay Divakaran, Regunathan Radhakrishnan, Kadir A. Peker USP 7,444,251. Filed: August 1, 2006. Granted: October 28, 2008. 12. Method and System for Segmenting Videos Using Face Detection Kadir A. Peker, Ajay Divakaran USP 7,555,149. Filed: October 25, 2005. Granted: June 30, 2009.