About me

I am a lecturer in Computer Science at the University of St Andrews.

I am interested in all aspects of computer vision, from low-level features to high-level reasoning and understanding. I am particularly interested in scene interpretation and modelling, and most of my work has focussed on middle-level processes and the transition from low-level features to semantic, interpretable entities. The combination of top-down and bottom-up processing typically plays an important role in my work.

I am also interested in applications of machine learning to real-world problems, such as applications of data science in education research, monitoring for conservation and healthcare applications, and data privacy.


CS5014 Machine Learning

CS3104 Operating Systems

CS1002 Object Oriented Programming

CS5003 Masters Programming projects (until 2018)

CS3101 Databases (until 2017)

CS1003 Programming with data (until 2017)


Contact me if you wish to discuss Senior Honours projects, MSc dissertations and PhD topics.

Shape detection with nearest neighbour contour fragments


We present a novel method for shape detection in natural scenes based on incomplete contour fragments and nearest neighbour search. In contrast to popular methods which employ sliding windows, chamfer matching and SVMs, we characterise each contour fragment by a local descriptor and perform a fast nearest-neighbour search to find similar fragments in the training set. Based on this idea, we show how to learn robust object models from training images, to generate reliable object hypotheses, and to verify them.

Publications: BMVC'15

BIMP: Fast, repeatable keypoints based on the visual cortex

GPU keypoints

Biologically inspired processing is often slow due to repeated convolutions. We have managed to create a biologically-inspired keypoint detection algorithm which is both fast and competitive with the state of the art in terms of repeatability. Efficient CPU and GPU-based implementations are available.

Publications: ICIP'13, Neurocomputing'15

Fast, learning free object detection based on Naive Bayes Nearest Neighbours


We designed a fast Bayesian algorithm for category-level object detection in natural images. We modified the popular Naive Bayes Nearest Neighbour classification algorithm to make it suitable for evaluating multiple sub-regions in an image, and offered a fast, filtering-based alternative to the multi-scale sliding window approach. Our algorithm is example-based and requires no learning. Tests on standard datasets and robotic scenarios show competitive detection rates and real-time performance of our algorithm. An implementation based on OpenCV is available.

Publications: ICIP'14

Context-based probabilistic scene interpretation

Facade segmentation

General scene understanding remains one of the great unsolved problems in Computer Vision and Artificial Intelligence. Building on deep Bayesian Compositional Hierarchies (BCH), we present a general framework for probabilistic scene understanding based on conceptual models and aggregate hierarchies. We show how uncertain evidence can be integrated into a stepwise interpretation process and how top-down context can resolve uncertainty in detection. Results are presented in the domain of facade images.

Publications: KI'08, UCVP'09, ICAART'10, IFIP AI'10, Thesis

Segmentation of 3D volumes

Fibre bundle

The aim of this project was to develop methods for analysing 3D tomography images of medium density fibreboards (MDF), a material made of wood commonly used in the industry. The tasks included segmentation of individual fibres, finding the contact surface between the fibres, determining the amount of adhesive resin, calculating the lumen volume and visualising the results.

Publications: SPIE'06, IVC'08, EPPS'06


See also: ORCID, DBLP, Google Scholar

G. Zemaityte and K. Terzić: Supervisor Recommendation Tool for Computer Science Projects, Proc. Int. Conf. on Computing Education Practice, Durham, Jan 2019 (accepted)

M. Saleiro, K. Terzić, J.M.F. Rodrigues and J.M.H. du Buf: BINK: Biological binary keypoint descriptor, Biosystems, Vol.162, pp. 147-156, 2017 [link]

K. Terzić, S. Krishna and J.M.H. du Buf: Texture features for object salience, Image and Vision Computing, Vol. 67, pp. 43-51, 2017 [link]

K. Terzić and J.M.H. du Buf: Interpretable Feature Maps for Robot Attention, Proc. HCI, Vancouver 2017, pp. 456-467, 2017 [pdf]

K. Terzić, M. Hansard: Methods for Reducing Visual Discomfort in Stereoscopic 3D: A Review, Signal Processing: Image Communication, Vol. 47, pp. 402-416, 2016 [link]

K. Terzić, S. Krishna and J.M.H. du Buf: A Parametric Spectral Model for Texture-Based Salience, Proc. GCPR, Aachen, pp. 331-342, Oct 2015 [pdf]

K. Terzić, H. L A. Mohammed and J.M.H. du Buf: Shape Detection with Nearest Neighbour Contour Fragments, Proc. BMVC, Swansea, Sep 2015 [bibtex] [pdf] [abstract]

M. Saleiro, M. Farrajota, K. Terzić, S. Krishna, J.M.F. Rodrigues and J.M.H. du Buf: Biologically inspired vision for human-robot interaction, Proc. HCI, Los Angeles, Aug 2015, pp. 505-517 [bibtex] [pdf]

K. Terzić, J.M.F. Rodrigues and J.M.H. du Buf: BIMP: A Real-Time Biological Model of Multi-Scale Keypoint Detection in V1, Neurocomputing, Vol. 150, pp. 227-237, Feb 2015 [bibtex] [preprint pdf] [link]

K. Terzić, D. Lobato, M. Saleiro and J.M.H. du Buf: A Fast Neural-Dynamical Approach to Scale-Invariant Object Detection, Proc. ICONIP, Kuching, LNCS vol. 8834, pp. 511-518, Nov 2014 [bibtex] [pdf]

M. Saleiro, K. Terzić, D. Lobato, J.M.F. Rodrigues and J.M.H. du Buf: Biologically Inspired Vision for Indoor Robot Navigation, Proc. ICIAR, Vilamoura, LNCS vol. 8815, pp. 469-477, Oct 2014 [bibtex] [pdf]

K. Terzić and J. M. H. du Buf: An Efficient Naive Bayes Approach to Category-Level Object Detection, Proc. IEEE ICIP, Paris, pp. 1658-1662, Oct 2014 [bibtex] [pdf]

O. Lomp, K. Terzić, C. Faubel, J. M. H. du Buf and G. Schöner: Instance-based Object Recognition with Simultaneous Pose Estimation Using Keypoint Maps and Neural Dynamics, Proc. ICANN, Hamburg, LNCS vol. 8681, pp. 451-458, Sep 2014 [bibtex] [pdf]

J.M.F. Rodrigues, K. Terzić, R. Lam and J.M.H. du Buf: Face and Object Recognition Using Biological Features and Few Views, in: Contemporary Advancements in Information Technology Development in Dynamic Environments, Mehdi Khosrow-Pour (ed.), IGI Global, pp. 58-77, 2014 [bibtex] [link]

K. Terzić, J.M.F. Rodrigues and J.M.H. du Buf: Fast Cortical Keypoints for Real-Time Object Recognition, Proc. IEEE ICIP, Melbourne, pp. 3372-3376, Sep 2013 [bibtex] [pdf]

K. Terzić, D. Lobato, M. Saleiro, J. Martins, M. Farrajota, J.M.F. Rodrigues and J.M.H. du Buf: Biological Models for Active Vision: Towards a Unified Architecture, Proc. ICVS, St. Petersburg, LNCS vol. 7963, pp. 113-122, Jul 2013 [bibtex] [pdf]

K. Terzić, J.M.F. Rodrigues and J.M.H. du Buf: Real-time Object Recognition Based on Cortical Multi-scale Keypoints, Proc. IbPRIA, Funchal, LNCS vol. 7887, pp. 314-321, Jun 2013 Best Paper Finalist [bibtex] [pdf]

M. Saleiro, M. Farrajota, K. Terzić, J.M.F. Rodrigues and J.M.H. du Buf: A Biological and Realtime Framework for Hand Gestures and Head Poses, Proc. HCI, Las Vegas, LNCS vol. 8009, pp 556-565 Jul 2013 [bibtex] [pdf]

J.M.H. du Buf, K. Terzić, J.M.F. Rodrigues: Phase-differencing in stereo vision - solving the localisation problem, Proc. BIOSIGNALS, pp. 254-263, Barcelona, Feb 2013 [bibtex] [pdf]

M. Farrajota, M. Saleiro, K. Terzić, J.M.F. Rodrigues and J.M.H. du Buf: Multi-scale cortical keypoints for realtime hand tracking and gesture recognition, Proc. Workshop on Cognitive Assistive Systems, IROS, pp. 9-15, Vilamoura, Oct 2012 [bibtex] [pdf]

K. Terzić: A Generic Middle Layer for Image Understanding, Doctoral Dissertation, University of Hamburg, 2011 [URL]

B. Neumann and K. Terzić: Context-based Probabilistic Scene Interpretation, Proc. IFIP AI, pp. 155-164, Brisbane, Sep 2010 [bibtex] [pdf]

K. Terzić, L. Hotz and J. Šochman: Interpreting Structures in Man-made Scenes; Combining Low-Level and High-Level Structure Sources, Proc. ICAART, Valencia, J. 2010 [bibtex] [pdf]

K. Terzić and B. Neumann: Integrating Context Priors into a Decision Tree Classification Scheme, Proc. MVIPPA, vol. 60, Bangkok, Dec 2009 [bibtex] [pdf]

A. Kreutzmann, K. Terzić and B. Neumann: Context-aware Classification for Incremental Scene Interpretation, Proc. Workshop on Use of Context in Vision Processing, UCVP, ICMI-MLMI, Boston, Nov 2009 [bibtex] [pdf]

K. Terzić and B. Neumann: Decision Trees for Probabilistic Top-down and Bottom-up Integration, Technical Report FBI-HH-B-288/09, University of Hamburg 2009 [bibtex] [pdf]

J. Hartz, L. Hotz, B. Neumann and K. Terzić: Automatic Incremental Model Learning for Scene Interpretation, Proc. IASTED CI, Honolulu, Aug 2009 [bibtex] [pdf]

L. Hotz, B. Neumann and K. Terzić: High-Level Expectations for Low-Level Image Processing, KI 2008: Advances in Artificial Intelligence, LNCS vol. 5243, pp. 87-94, 2008 [bibtex] [pdf]

P. Stelldinger and K. Terzić: Digitization of Non-regular Shapes in Arbitrary Dimensions, in: Image and Vision Computing Volume 26, Issue 10, pp. 1338-1346, Oct 2008 [bibtex] [preprint pdf] [link]

K. Terzić, L. Hotz and B. Neumann: Division of Work During Behaviour Recognition - The SCENIC Approach, Proc. Behaviour Monitoring and Interpretation Workshop, KI, Bremen, Sep 2007 [bibtex] [pdf]

L. Hotz, B. Neumann, K. Terzić and J. Šochman: Feedback between Low-level and High-level Image Processing, Technical Report FBI-HH-B-278/07. University of Hamburg 2007 [bibtex] [pdf]

T. Walther, H. Thoemen, K. Terzić and H. Meine: New Opportunities for the Microstructural Analysis of Wood Fiber Networks, Proc. EPPS, Llandudno, 2006, pp. 23--32 [bibtex] [pdf]

T. Walther, K. Terzić, T. Donath, H. Meine, F. Beckmann and H. Thoemen: Microstructural analysis of lignocellulosic fiber networks, Developments in X-Ray Tomography V, vol. 6318. San Diego, 2006, CID number 631812 [bibtex] [pdf]

K. Terzić: Evaluating Architectural Features of Programmable Microprocessors, Master Thesis, Arbeitsbereich Technische Informatik VI, Technische Universität Hamburg-Harburg, 2003