Selected Publications

In this paper we explore how deep convolutional neural networks dedicated to the task of object detection can improve our industrial-oriented object pipelines. We do this by using single pass (both classification and detection) deep learned architectures in combination with transfer learning to drastically drop the amount of necessary training data. Furthermore we prove the efficiency of these algorithms in both single- and multi-class scenarios.
VISAPP, 2018

In this work we prove that using cascade classifiers yields promising results on coconut tree detection in aerial images. However with the rise of robust deep learning algorithms for both detection and classification, and the significant drop in hardware costs, we wonder if it is feasible to apply deep learning to solve the task of fast and robust coconut tree localization in aerial imagery. By exploring deep learning we prove that it is indeed a feasible alternative for coconut tree detection with a high average precision.
VISAPP, 2018

In this PhD dissertation we explore the possibility to use scene-specific constraints of industrial object detection tasks to influence three main aspects of object detection algorithms: the amount of training data, the speed of the detection process and the amount of FP/FN detections. Moreover, we propose steps to simplify the training process using techniques like active learning and data augmentation.
Doctoral Manuscript,2017

Recent Publications

. Automated visual fruit detection for harvest estimation and robotic harvesting. IPTA, 2016.

PDF Slides

. Detection of photovoltaic installations in RGB aerial imaging: a comparative study. GEOBIA, 2016.

PDF Slides

. Safeguarding Privacy by Reliable Automatic Blurring of Faces in Mobile Mapping Images. VISAPP, 2016.

PDF Slides

. OpenCV 3 Blueprints: Expand your knowledge of computer vision by building amazing projects with OpenCV 3. PACKT Publishing, 2015.

PDF Code

. Automated Walking Aid Detector Based on Indoor Video Recordings. EMBC, 2015.

PDF Poster

Recent & Upcoming Talks

Science Figured Out
Wed, Jun 6, 2018 10:30
Sound Of Science Festival
Sun, May 27, 2018 10:30
Finale van de Wetenschapsbattle 2018
Fri, Mar 16, 2018 09:00
Wetenschapsbattle 2018 - Voorronde provincie Antwerpen
Mon, Mar 12, 2018 09:00
Public PhD Defense
Wed, Dec 13, 2017 17:30
Deep Learning Workshop
Wed, May 10, 2017 08:30

Teaching

Academic courses

  • Research methodologies | 2018 - now

  • Scientific seminars | 2017 - now

    List of seminars: https://iiw.kuleuven.be/onderzoek/eavise/seminars

  • Computer vision hands-on labs using the OpenCV framework | 2015 - now

  • Scientific project for second year students | 2014 - 2017

    Marker based augmented reality | 2017 & 2016

    Mini-me in 3D (creating + printing 3D models of students) | 2016

    Building an anti burglar system using sensor fusion | 2014 & 2015

    Building a laser line based 3D scanner | 2015

    360 degree panoramic photo helm | 2014

Master thesisses

  • Automated coconut tree detection in aerial imagery.

  • Improved the existing orchid flower detection pipeline.

  • Man overboard detection using thermal cameras.

  • Visual framework for keypoint descriptors and matchers in OpenCV3.

  • Global framework for defining the orientation of images patches.

  • Detection, classification and tracking of trafic at crossroads.

  • Detection and classification of cars and pedestrians in IR images.

  • Efficient person detection in video feeds.

Contact