Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision		Previous revision
Next revisionBoth sides next revision
teaching:gsoc2018 [2018/01/21 20:29] – balintbeteaching:gsoc2018 [2018/01/22 09:36] – [openEASE -- Web-based Robot Knowledge Service] ahaidu
Line 38: Line 38:
 RoboSherlock builds on top of the ROS ecosystem and is able to wrap almost any existing perception algorithm/framework, and allows easy and coherent combination of the results of these. The framework has a close integration with two of the most popular libraries used in robotic perception, namely OpneCV and PCL. More details about RoboSherlock can be found on the project [[http://robosherlock.org/|webpage]]. RoboSherlock builds on top of the ROS ecosystem and is able to wrap almost any existing perception algorithm/framework, and allows easy and coherent combination of the results of these. The framework has a close integration with two of the most popular libraries used in robotic perception, namely OpneCV and PCL. More details about RoboSherlock can be found on the project [[http://robosherlock.org/|webpage]].
  
 +===== openEASE -- Web-based Robot Knowledge Service =====
  
 +OpenEASE is a generic knowledge database for collecting and analyzing experiment data. Its foundation is the KnowRob knowledge processing system and ROS, enhanced by reasoning mechanisms and a web interface developed for inspecting comprehensive experiment logs. These logs can be recorded for example from complex CRAM plan executions, virtual reality experiments, or human tracking systems. OpenEASE offers interfaces for both, human researchers that want to visually inspect what has happened during a robot experiment, and robots that want to reason about previous task executions in order to improve their behavior.
 +
 +The OpenEASE web interface as well as further information and publication material can be accessed through its publicly available [[http://www.open-ease.org/|website]]. It is meant to make complex experiment data available to research fields adjacent to robotics, and to foster an intuition about robot experience data.
 +
 +===== RobCoG - **Rob**ot **Co**mmonsense **G**ames =====
 +
 +[[http://robcog.org/|RobCoG]] is a framework consisting of various open source games and plugins written in the Unreal Engine with the intention of collecting and equipping robots with commonsense and naive physics knowledge.
 ===== Proposed Topics ===== ===== Proposed Topics =====
  
Line 76: Line 84:
  
  
-==== Topic 1: Multi-modal Cluttered Scene Analysis in Knowledge Intensive Scenarios ====+==== Topic 2: Felxible perception pipeline manipulation for RoboSherlock ====
  
 {{  :teaching:gsoc:topic1_rs.png?nolink&200|}} {{  :teaching:gsoc:topic1_rs.png?nolink&200|}}
  
-**Main Objective:** In this topic we will develop algorithms that en- +**Main Objective:** RoboSherlock is based on the unstructured information management paradigm and uses the uima library at it's core. The c++ implementation of this library is limited multiple ways. In this topic you will develop a module in order to flexibly manage perception pipelines by extending the current implementation to enable new modalities and  run pipelines in parallel. This involves implementing an API for pipeline and data handling that is rooted in the domain of UIMA. 
-able robots in a human environment to recognize objects in diffi- +
-cult and challenging scenarios. To achieve this the participant will +
-develop annotators for RoboSherlock that are particularly aimed at +
-object-hypotheses generation and merging. Generating a hypotheses +
-essentially means to generate regions/clusters in our raw data that +
-form a single object or object-part. In particular this entails the de- +
-velopment of segmentation algorithms for visually challenging scenes +
-or object properties, as the likes of transparent objects, or cluttered, +
-occluded scenes. The addressed scenarios include stacked, occluded +
-objects placed on shelves, objects in drawers, refrigerators, dishwash- +
-ers, cupboards etc. In typical scenarios, these confined spaces also +
-bare an underlying structure, which will be exploited, and used as +
-background knowledge, to aid perception (e.g. stacked plates would +
-show up as parallel lines using an edge detection). Specifically we +
-would start from (but not necessarly limit ourselves to) the implemen- +
-tation of two state-of-the-art algorithms described in recent papers: +
- +
-[1] Aleksandrs Ecins, Cornelia Fermuller and Yiannis Aloimonos, Cluttered Scene Segmentation Using the Symmetry Constraint, International Conference on Robotics and Automation(ICRA) 2016 +
-[2] Richtsfeld A., M ̈ +
-orwald T., Prankl J., Zillich M. and Vincze +
-M. - Segmentation of Unknown Objects in Indoor Environments. +
-IEEE/RSJ International Conference on Intelligent Robots and Sys- +
-tems (IROS), 2012.+
  
-**Task Difficulty:** The task is considered to be challenging, as it is still a hot research topic where general solutions do not exist.+**Task Difficulty:** The task is considered to be of medium difficulty. 
      
 **Requirements:** Good programming skills in C++ and basic knowledge of CMake and ROS. Experience with PCL, OpenCV is prefered. **Requirements:** Good programming skills in C++ and basic knowledge of CMake and ROS. Experience with PCL, OpenCV is prefered.
  
-**Expected Results:** Currently the RoboSherlock framework lacks good perception algorithms that can generate object-hypotheses in challenging scenarios(clutter and/or occlusion). The expected results are several software components based on recent advances in cluttered scene analysis that are able to successfully recognized objects in the scenarios mentioned in the objectives, or a subset of these.+**Expected Results:** an extension to RoboShelrock that allows splitting and joingin pipelines, executing them in parallel, merging results from multiple types of cameras etc. 
  
 Contact: [[team/ferenc_balint-benczedi|Ferenc Bálint-Benczédi]] Contact: [[team/ferenc_balint-benczedi|Ferenc Bálint-Benczédi]]
  
  
- Impressum - Datenschutz



Prof. Dr. hc. Michael Beetz PhD
Head of Institute

Contact via
Andrea Cowley
assistant to Prof. Beetz
ai-office@cs.uni-bremen.de

Discover our VRB for innovative and interactive research


Memberships and associations:


Social Media: