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
jobs [2014/07/11 11:50] – [Theses and Jobs] jworchjobs [2017/01/24 07:19] – [Theses and Jobs] bartelsg
Line 3: Line 3:
 If you are looking for a bachelor/master thesis or a job as a student research assistant, you may find some interesting opportunities on this page. If you are looking for a bachelor/master thesis or a job as a student research assistant, you may find some interesting opportunities on this page.
  
 +<html><!--
 +== Lisp / CRAM support assistant (HiWi) ==
  
 +Technical support for the group for Lisp and the CRAM framework. \\
 +5 hours per week for up to 1 year (paid).
  
-== GPU-based Parallelization of Numerical Optimization Techniques (BA/MA/HiWi)==+Requirements: 
 +  * Good programming skills in Common Lisp 
 +  * Basic ROS knowledge
  
-In the field of Machine Learning, numerical optimization techniques play a focal role. However, as models grow larger, traditional implementations on single-core CPUs suffer from sequential execution causing a severe slow-down. In this thesis, state-of-the-art GPU frameworks (e.g. CUDA) are to be investigated in order implement numerical optimizers that substantially profit from parallel execution.+The student will be introduced to the CRAM framework at the beginning of the job, which is a robot programming framework written in Lisp. The student will then be responsible for assisting not familiar with the framework people, explaining them the parts they don't understand and pointing them to the relevant documentation sources.
  
-Requirements: +Contact: [[team:gayane_kazhoyan|Gayane Kazhoyan]] 
-  * Skills in numerical optimization algorithms +--> 
-  * Good programming skills in Python and C/C+++</html>
  
-Contact: [[team:daniel_nyga|Daniel Nyga]] 
  
-== Online Learning of Markov Logic Networks for Natural-Language Understanding (MA)==+== Integrating PR2 in the Unreal Game Engine Framework (BA)== 
 + {{ :research:unreal_ros_pr2.png?200|}} 
  
-Markov Logic Networks (MLNs) combine the expressive power of first-order logic and probabilistic graphical models. In the past, they have been successfully applied to the problem of semantically interpreting and completing natural-language instructions from the web. State-of-the-art learning techniques mostly operate in batch mode, i.e. all training instances need to be known in the beginning of the learning process. In context of this thesis, online learning methods for MLNs are to be investigated, which allow incremental learning, when new examples come in one-by-one.+Integrating the [[https://www.willowgarage.com/pages/pr2/overview|PR2]] robot with [[http://www.ros.org/|ROS]] support in the [[https://www.unrealengine.com|Unreal Engine 4]] Framework.
  
 Requirements: Requirements:
-  * Experience in Machine Learning. +  * Good programming skills in C/C++ 
-  * Experience with statistical relational learning (e.g. MLNs) is helpful. +  * Basic physics/rendering engine knowledge 
-  * Good programming skills in Python.+  * Basic ROS knowledge 
 +  * UE4 basic tutorials
  
-Contact: [[team:daniel_nyga|Daniel Nyga]]+Contact: [[team:andrei_haidu|Andrei Haidu]]
  
  
-==HiWi-Position: Knowledge Representation & Language Understanding for Intelligent Robots==+== Realistic Grasping using Unreal Engine (BA/MA) ==
  
-In the context of the European research project RoboHow.Cog [1,2] we +{{  :teaching:gsoc:topic2_unreal.png?nolink&150|}}
-are investigating methods for combining multimodal sources of knowledge (e.g. video, natural-language recipes or computer games), in order to enable mobile robots to autonomously acquire new high level skills like cooking meals or straightening up rooms.  +
  
-The Institute for Artificial Intelligence is hiring a student researcher for the +The objective of the project is to implement var- 
-development and the integration of probabilistic methods in AI, which enable intelligent robots to understand, interpret and execute natural-language instructions from recipes from the World Wide Web.+ious human-like grasping approaches in a game developed using [[https://www.unrealengine.com/|Unreal Engine]]. 
  
-This HiWi-Position can serve as a starting point for future Bachelor's, Master's or Diploma Theses.+The game consist of a household environment where a user has to execute various given tasks, such as cooking a dish, setting the table, cleaning the dishes etc. The interaction is done using various sensors to map the users hands onto the virtual hands in the game.
  
-Tasks: +In order to improve the ease of manipulating objects the user should 
-  * Implementation of an interface to the Robot Operating System (ROS). +be able to switch during runtime the type of grasp (pinch, power 
-  * Linkage of the knowledge base to the executive of the robot. +grasp, precision grip etc.) he/she would like to use. 
-  * Support for the scientific staff in extending and integrating components onto the robot platform PR2.+   
 +Requirements:  
 +  * Good programming skills in C++ 
 +  * Good knowledge of the Unreal Engine API.  
 +  * Experience with skeletal control / animations / 3D models in Unreal Engine. 
 + 
 + 
 +Contact: [[team/andrei_haidu|Andrei Haidu]] 
 + 
 +== Kitchen Activity Games in a Realistic Robotic Simulator (BA/MA)== 
 + {{ :research:gz_env1.png?200|}}  
 + 
 +Developing new activities and improving the current simulation framework done under the [[http://gazebosim.org/|Gazebo]] robotic simulator. Creating a custom GUI for the game, in order to launch new scenarios, save logs etc.
  
 Requirements: Requirements:
-  * Studies in Computer Science (Bachelor's, Master's or Diploma) +  * Good programming skills in C/C++ 
-  * Basic skills in Artificial Intelligence +  * Basic physics/rendering engine knowledge 
-  * Optional: basic skills in Probability Theory +  * Gazebo simulator basic tutorials
-  * Optional: basic skills in Machine Learning +
-  * Good programming skills in Python and Java+
  
-Hours: 10-20 h/week+Contact: [[team:andrei_haidu|Andrei Haidu]]
  
-Contact: [[team:daniel_nyga|Daniel Nyga]] 
  
-[1] www.robohow.eu\\ 
-[2] http://www.youtube.com/watch?v=0eIryyzlRwA 
  
  
-== Depth-Adaptive Superpixels (BA/MA)== +== Automated sensor calibration toolkit (BA/MA)==
- {{ :research:dt_dasp.png?200|}} +
-We are currently investigating a new set of sensors (RGB-D-T), which is a combination of a kinect with a thermal image camera. Within this project we want to enhance the Depth-Adaptive Superpixels (DASP) to make use of the thermal sensor data. Depth-Adaptive Superpixels oversegment an image taking into account the depth value of each pixel.+
  
-Since the current implementation of DASP is not very performant for high resolution images, there are several options for doing a project in this field like reimplementing DASP using CUDA, investigating how thermal data can be integrated, ...+Computer vision is an important part of autonomous robots. For robots the image sensors are the main source of information of the surrounding world. Each camera is different, even if they are from the same production line. For computer vision, especially for robots manipulating their environment, it is important that the parameters for the cameras in use are well known. The calibration of a camera is a time consuming task, and the result depends highly on the chosen setup and the accuracy of the operator.
  
-Requirements: +The topic for this thesis is to develop an automated system for calibrating cameras, especially RGB-D cameras like the Kinect v2.
-  * Basic knowledge of image processing +
-  * Good programming skills in C/C++. +
-  * Experience with CUDA is helpful+
  
-Contact: [[team:jan-hendrik_worch|Jan-Hendrik Worch]]+ {{ :kinect2_calibration_setup_small.jpg?200|}} 
 +The system should: 
 +  * be independent of the camera type 
 +  * estimate intrinsic and extrinsic parameters 
 +  * calibrate depth images (case of RGB-D) 
 +  * integrate capabilities from Halcon [1] 
 +  * operate autonomously
  
 +Requirements: 
 +  * Good programming skills in Python and C/C++
 +  * ROS, OpenCV
  
-== Depth-Adaptive Superpixels (BA/MA)== +[1] http://www.halcon.de/
- {{ :research:human_model.png?200|}}+
  
-For tracking people, the use of particle filters is a common approach. However, the quality of those filters heavily depends on the way particles are spread. In this thesis, a library for the physical simulation of a human model is to be implemented.+Contact: [[team:alexis_maldonado|Alexis Maldonado]] and [[team:thiemo_wiedemeyer|Thiemo Wiedemeyer]]
  
-Requirements:+/* 
 +== On-the-fly 3D CAD model creation (MA)== 
 + 
 +Create models during runtime for unknown textured objets based on depth and color information. Track the object and update the model with more detailed information, completing it's 3D model from multiple views improving redetection. Using the robots manipulator pick up the object and complete the model by viewing it from multiple viewpoints. 
 + 
 +Requirements: 
   * Good programming skills in C/C++   * Good programming skills in C/C++
-  * Optional: Experience in working with physics libraries such as Bullet+  * strong background in computer vision  
 +  * ROS, OpenCV, PCL
  
-Contact: [[team:jan-hendrik_worch|Jan-Hendrik Worch]]+Contact: [[team:thiemo_wiedemeyer|Thiemo Wiedemeyer]]
  
 +== Simulation of a robots belief state to support perception(MA) ==
 +
 +Create a simulation environment that represents the robots current belief state and can be updated frequently. Use off-screen rendering to investigate the affordances these objects possess, in order to support segmentation, detection and tracking of these in the real world. 
 +
 +Requirements: 
 +  * Good programming skills in C/C++
 +  * strong background in computer vision 
 +  * Gazebo, OpenCV, PCL
 +
 +Contact: [[team:ferenc_balint-benczedi|Ferenc Balint-Benczedi]]
 +*/
 +
 +== Multi-expert segmentation of cluttered and occluded scenes ==
 +
 +Objects in a human environment are usually found in challenging scenes. They can be stacked upon eachother, touching or occluding, can be found in drawers, cupboards, refrigerators and so on. A personal robot assistant in order to execute a task, needs to detect these objects and recognize them. In this thesis a multi-modal approach to interpreting cluttered scenes is going to be investigated, combining the results of multiple segmentation algorithms in order to come up with more reliable object hypotheses.
 +
 +Requirements: 
 +  * Good programming skills in C/C++
 +  * strong background in 3D vision 
 +  * basic knowledge of ROS, OpenCV, PCL
  
 +Contact: [[team:ferenc_balint-benczedi|Ferenc Balint-Benczedi]]
- 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: