We offer students a summer job in the Robotic Perception Group (ROP) and the Applied Algebra and Geometry Group (AAG). Students become student research assistants working on a particular research project and paid for by the project. Each student will have a supervisor. The specific agreement has to be done with her or him. See the list of topics below. Contact the supervisor of the topic you like.
The usual student summer job lasts from 4 to 6 weeks. The job can be divided into several non-consecutive periods of time. The first two weeks serve for testing. If the student does not like the job or if the supervisor is not happy with the student’s performance, the job can be stopped after two weeks. Students who have not studied at the university yet are offered jobs for two weeks initially. If both sides of the contract, the student and the supervisor like it, the job can be prolonged.
The pay is counted for 160 hours per months at the following rates (before the tax is paid by the student) in CZK per hour → CZK per month: after A-level from the secondary school 130 → 20,800; after Bc1 140 → 22,400; after Bc2 150 → 24,000; after Bc3 160 → 25,600, after Mag1 170 → 27,200, after Mag2 180 → 28,800, after Mag/Ing graduation 200 → 30,400. The median of gross wages in the Czech Republic was CZK 32,870 in the Q4 of the year 2020.
The list of topics, their brief descriptions, and supervisors follow. Please, contact the supervisor directly if you are interested in a particular free topic. If you need more general information or you do not know which topic to select, call Václav Hlaváč on his mobile phone number +420 603 149 689. He will help you. You can call between 7:00 and 20:00.
If the student passed the entry interview with Václav Hlaváč or worked with an advisor before she/he appears at the managerial table. Everyone who knows the link can see it. (Note: V. Hlaváč had to remove phone numbers and email addresses from the managerial table. It was misused for unwanted phone calls. Advisors have access to the table with phone numbers and emails.) Prospective advisors can also edit it. The table helps advisors to find a student for the project. If there was a preference of a student for a particular topic, I mentioned it in the table. When the pairing is agreed upon, the advisor mentions it in the table in the column Interested advisor. I will do the final assignment. Later, the table will help our administrative staff (Petra Žufníčková and Zuzana Kadavá) to make contracts with students. I plan to create a GANNT chart enabling me to synchronize better between students and their advisors.
- Topic A: Semantic representation of a traffic scene and qualitative reasoning in it, e.g. for self-driving car
Supervisor: Václav Hlaváč
Description: We will work with OpenStreetMap, add objects to it and build a simple logic-based reasoning system.
Team: Petr Ježek
Note: a senior student (Petr Ježek) and one junior student (to be found), knowledge in Artificial Intelligence is an advantage - Topic B: Contribution to a telemedicine software allowing a remote observation of, e.g. artificial respiratory ventilator (Czech umělá plicní ventilace)
Supervisors: Lenka Vysloužilová
Description: Part of the project VENT-CONNECT. There is a cheap device connected to the artificial respiratory ventilator and software running on a physician notebook computer at a remote place. The student will contribute to rewriting a prototype software in the physician notebook.
Team: Richard Fritsch, Roman Novikov
Note: The development is in Kotlin programming language. It is enough if a student has experience with Java or C#. - Topic C: Testing and development of human tracking software in surveillance videos from retail shops
Supervisors: Václav Hlaváč, Radoslav Škoviera
Description: We contribute to a probabilistic tracker, i.e. program able to detect and track people. The student team will test the tracker on videos, help to annotate videos used for machine learning of the detector/tracker.
Team: Tereza Uhrová, One to three students
Note: The development will be in Python. Its knowledge is desirable. - Topic D: Tactile feedback on a force/torque compliant robot
Supervisor: Václav Hlaváč, Libor Wagner
Description: The aim is to create an eye-catching demo demonstrating a peg-in-hole assembly task using Robotiq gripper and Franka Emika Panda robot.
Team: Two students: Lucie Vajnerová plus one more student
Note: The development will be in Python and ROS2. - Topic E: Robot calibration
Supervisor: Pavel Krsek
Description: The robot is manufactured and mounted with limited accuracy. Using an external LEICA laser rangefinder, we can measure its position and obtain the position of the manipulator or estimate some parameters of kinematics.
Team: Vratislav Besta plus one more student
Note: Python, partially ROS - Topic F: Manual positioning of sensors
Supervisor: Pavel Krsek
Description: For the purposes of the experimental workplace, it is necessary to adjust the position of the sensors. The position will be measured by a LEICA laser rangefinder. The goal is to provide an automatic system to guide the operator during setup.
Team: 1 student
Note: None - Topic G: Motorized sensors holder
Supervisor: Pavel Krsek
Description: Design a motor-driven holder (kinematics). In the future, implementation of the controller and interface in a computer. First, the 3D drawings and computations must be done.
Team: 1 student
Note: 3D design software, Python - Topic H: Focus of the lens
Supervisor: Pavel Krsek
Description: We need to solve the task of focusing the image of the „lens“. The lens appears as a discontinuous circle. In the sequence of images with a different focus, the one where the lens image is the sharpest.
Team: 1 student
Note: None - Topic I: Preparation of teaching materials for robotics
Supervisor: Vladimír Smutný, Pavel Krsek
Description: Preparing images (3D models), videos, interactive models, and other materials for teaching the subject of Robotics.
Team: 1 student
Note: The choice of tools is very flexible, but the results should be easily edited and shared with students as well. - Topic J: Robot Hermes
Supervisor: Vladimír Smutný
Description: For the purpose of transporting patients within the rehabilitation facility, it is necessary to develop or modify the existing software for navigation and planning of an autonomous mobile robot.
Team: 1 – 3 students
Note: Python, ROS, Matlab - Topic K: Manipulation robotic tasks in myGym environment
Supervisor: Michal Vavrečka
Description: Train industrial robots in virtual environment to be able to manipulate with objects when moving distractors are present
Team: Jonáš Križ, Vitězslav Žáček
Note: Python, ML - Topic L: Recommendation chatbot for Pepper robot
Supervisor: Michal Vavrečka
Description: Implementation of the knowledge base and topics switcher for the recommendation chatbot
Team: 1 student
Note: Python
- Topic M: Applications for the Pepper humanoid robot
Supervisor: Michal Vavrečka
Description: The applications will serve the HUMR project
Team: 1 student
Note: Python
- Topic N: Trajectory planning for a robot with 9 DOF
Supervisor: Vladimír Smutný
Description: Robotic cell has 9 DOF while the task itself has 6 DOF. The final goal is to convert CAD/CAM output to the optimal trajectory of the robot.
Team: 2 students
Note: One subtask is more programming oriented, the second one focuses on trajectory optimization and thus mathematics is emphasized. Implementation is in ROS.
- Topic O: Automatic road lines annotation from aerial/satellite images
Supervisor: Miroslav Uller
Description: The aim is to help to annotate data captured by self-driving car sensors by external information from aerial/satellite images. This should improve the reliability of annotations.
Team: 2 students
Note: Familiarity with Python expected. Experience with Open CV, Anaconda, basic UNIX shell programming is a bonus. - Topic P: Automatic text processing in job announcement for statistical analysis
Supervisor: Miroslav Uller
Description: Students will explore web scraping and Silenium to analyze a free text. They will seek key phrases.
Team: 1 student
Note: Python basic knowledge expected. Unix shell, basic commands.
- Topic Q: Design of robot for carbon fiber laying
Supervisor: Vladimír Smutný
Description:
Team: Daniel Král
Note: None - Topic R: GUI for robotic workspace/ROS2 visualization
Supervisor: Radoslav Škoviera
Description: Improving and maintaining a user interface (currently partially implemented) in wxPython framework. The GUI is used in a collaborative robotic workspace. It will collect and visualize data from a database and ROS2.
Team: 1 (maybe 2) student(s)
Note: Python, wxPython, ROS 2 (using Python)
- Topic S: Learning manipulation skills from videos
Supervisor: Vladimir Petrik, Kateryna Zorina, Jiri Sedlar, Josef Sivic
Description: Finding and collecting videos of person-object interactions and annotation of contacts/objects. We use the dataset for learning and evaluation of 3D motion reconstruction of person-object interactions from videos, incl. contacts and forces. For more advanced students with interest and some prior experience in machine learning and visual recognition, the project could also involve training and evaluating object/contact recognizers and extending the state-of-the-art method to additional tools and skills. Please see: https://www.ciirc.cvut.cz/vysledek-ciirc-cvut-se-dostal-do-uzsiho-vyberu-nejlepsich-clanku-prestizni-konference-cvpr-v-pocitacovem-videni/
Team:
Note: None for collecting videos and annotation. Python for training and evaluation. - Topic T: Analyzing image samples of soil contamination
Supervisor: Martin Matoušek
Description: Cooperation with the company Dekonta
Team: 1 – 2 students, beginners are fine
Note: None - Topic U:
Supervisor:
Description:
Team:
Note: None
Responsible: Václav Hlaváč
