About
The Visual Cognitive Systems Laboratory is involved in basic and applied research of visually enabled intelligent systems addressing various research problems from the fields of computer vision, (deep) machine learning, and cognitive robotics.
The main research tasks include visual object tracking, detection, categorization, and segmentation, applied to various applications such as visual inspection for quality control, visual surveillance, and robot navigation.
Location
The laboratory is located at the Faculty of Computer and Information Science building, at the outskirts of Ljubljana. Detailed information on how to get there can be found here.
News
-
September 2024Center Direction Network for Grasping Point Localization on Cloths paper that introduces ViCoS Towel Dataset and a winning method from ICRA 2023 Cloth Competiton has been published in IEEE Robotics and Automation Letters.
-
May 2024ViCoS Team achived second place in the ICRA 2024 Cloth Competition challange, outcompeting nine other teams in cloth manipulation subtrack of the 9th Robotic Grasping and Manipulation Competition
-
May 2024Dense Center-Direction Regression for Object Counting and Localization with Point Supervision has been accepted to Pattern Recognition journal
-
February 2024A New Dataset and a Distractor-Aware Architecture for Transparent Object Tracking has been accepted to International Journal of Computer Vision
-
September 2023Automated detection and segmentation of cracks in concrete surfaces using joined segmentation and classification deep neural network paper has been accepted for publication in Construction and Building Materials journal.
-
June 2023ViCoS Team has won the perception challange on 2th Cloth and Manipulation Challenge that was part of 7th Robotic Grasping and Manipulation Competition of ICRA 2023
-
May 2023Domen Tabernik, a member of ViCoS, has been awarded by the Slovenian Pattern Recognition Society for his outstanding scientific article titled Segmentation-Based Deep-Learning Approach for Surface-Defect Detection
-
November 2022The paper Trans2k: Unlocking the Power of Deep Models for Transparent Object Tracking received the Best Paper Prize at the BMVC 2022
-
September 2022Trans2k: Unlocking the Power of Deep Models for Transparent Object Tracking accepted to the BMVC 2022 conference as an oral presentation
-
December 2021A Discriminative Single-Shot Segmentation Network for Visual Object Tracking accepted to IEEE TPAMI