André Kestler
AI Research Assistant
About Me
Hello, my name is André Kestler. I am currently working at Ostbayerische Technische Hochschule Amberg-Weiden (OTH-AW) as a research asstistant in the field of artificial intelligence. I completed my Master's degree in Artificial Intelligence (M.Sc.) and my Bachelor's degree in Electrical Engineering and Information Technology (B.Eng.), specialising in automation technology, at OTH-AW. I wrote my master's thesis in the field of non-intrusive load monitoring (NILM) and my bachelor's thesis in the field of computer vision in robotics.
My hobbies are road cycling, running and bouldering.
Bio
Age
26
Email
kestler.andre.code@gmail.com
Phone
+49 151 58811786
Address
Am Schützenheim 2, 92237 Sulzbach-Rosenberg, Deutschland
Skills
Languages
C
C++
Python
HTML
CSS
JavaScript
Python Libraries
Numpy
Matplotlib
Pandas
PyTorch
OpenCV
Experience
Working student Siemens AG in Amberg
Dez, 2021 - Okt, 2023
Practical semester Baumann GmbH in Amberg
Aug, 2019 - Dez, 2019
Specialised practical training Technical college in Amberg
Sept, 2015 - Juli, 2016
Metal and electrical apprenticeships
Education
University of Applied Sciences Ostbayerische Technische Hochschule Amberg-Weiden
2017 - 2024
2021 B.Eng., Electrical engineering and information technology
2024 M.Sc., Artificial Intelligence
2024 M.Sc., Artificial Intelligence
Technical college Berufliches Schulzentrum Amberg
2014 - 2017
2014/2015 Pre-class
2017 Advanced technical college entrance qualification
2017 Advanced technical college entrance qualification
Secondary school Krötensee-Mittelschule Sulzbach-Rosenberg
2008 - 2014
2013 Qualifying secondary school leaving certificate
2014 Intermediate qualification
2014 Intermediate qualification
Primary school Jahnschule Sulzbach-Rosenberg
2004 - 2008
Portfolio
Smart Mirror Project
2019 - 2021
A smart bathroom mirror that is controlled using gestures and has features such as face and emotion recognition. With the help of a camera can also be used to check medication intake.
Object recognition with computer vision and YOLOv3 Project
2020
The YOLOv3 neural network, together with a set of self-recorded data, was used to train a system to recognise a variety of tools and components in a laboratory setting. The data set for training was created and labelled in-house.
Pose estimation of ArUco markers using 2D and 3D image processing systems Bachelor thesis
2020 - 2021
A camera and markers were used to determine the position and orientation of components in a robot cell. The aim of the work was for the robot arm to be able to grip the parts without having to teach the points directly. In particular, the mobile robot should be able to grip the components even after changing position.
Natural Language Processing - Search engine Student research project
2021
A semantic search engine was developed as part of a student research project, with the objective of providing a tool that would facilitate the analysis of legal texts. The project was trained with court texts from openlegaldata.io. The user interface was designed to provide guidance and support during the search process. The search is supported by n-grams and word embeddings.
AR/VR - Hedgehog simulator Student research project
Robot Mapping with ROS Student research project
Frontier Exploration with ROS Student research project
Big Data and Cloud Computing - Computer Vision Pipeline Student research project
2023
In the Big Data and Cloud Computing course, a computer vision pipeline was developed as a student research project. The project consists of a frontend container and a backend container. The frontend is based on React and the backend on Python. Data is exchanged via an Amazon S3 bucket and the application is deployed in an Amazon EC2 instance. The user can upload an image and compile a pipeline consisting of different image processing steps. The pipeline is then applied to the image and the result is displayed to the user.
Deep Vision - YOLO Networks Student research project
2023
As part of the Deep Vision lecture, students are required to complete a project in the area of the course. The work will be done individually. As a project, two YOLO (You Only Look Once) networks are trained with the Udacity Self Driving Car data set and compared with each other. YOLOX and YOLOv8 are used in the following work.
Non-Intrusive Load Monitoring (NILM) - Load Disaggregation for a Production Module Master thesis
2023 - 2024
In this master's thesis, the energy consumption of a production module in an assembly line for contactors was analysed in detail. The work comprises several phases, including data acquisition, data analysis, event detection, clustering of the detected events and load disaggregation to the individual consumers. A separate data set, which was labelled using a video, serves as the basis for the clustering. The clusters of the labelled data are compared with the clusters of the unknown data in order to assign the most similar class (consumer).
Imprint
André Kestler
Am Schützenheim, 2
92237 Sulzbach-Rosenberg
Phone: +49 151 58811786
E-Mail: Contact
Disclaimer - legal information
§ 1 Warning regarding contentThe free and freely accessible content of this website has been created with the greatest possible care. However, the provider of this website assumes no liability for the accuracy and timeliness of the free and freely accessible journalistic guides and news provided. Contributions labelled by name reflect the opinion of the respective author and not always the opinion of the provider. No contractual relationship is established between the user and the provider simply by accessing the free and freely accessible content; in this respect, the provider's intention to be legally bound is lacking.
§ 2 External links
This website contains links to third-party websites (‘external links’). These websites are subject to the liability of the respective operators. When the external links were first created, the provider checked the third-party content for any legal violations. No legal offences were apparent at that time. The provider has no influence whatsoever on the current and future design and content of the linked pages. The inclusion of external links does not mean that the provider adopts the content behind the reference or link as its own. It is not reasonable for the provider to constantly monitor external links without concrete evidence of legal violations. However, such external links will be deleted immediately if the provider becomes aware of any legal violations.
§ 3 Copyright and neighbouring rights
The content published on this website is subject to German copyright and ancillary copyright law. Any use not authorised by German copyright and ancillary copyright law requires the prior written consent of the provider or respective rights holder. This applies in particular to the duplication, editing, translation, storage, processing or reproduction of content in databases or other electronic media and systems. Third-party content and rights are labelled as such. The unauthorised reproduction or distribution of individual contents or complete pages is not permitted and is punishable by law. Only the production of copies and downloads for personal, private and non-commercial use is permitted.
The presentation of this website in external frames is only permitted with written authorisation.
§ 4 Special terms of use
Insofar as special conditions for individual uses of this website deviate from the aforementioned paragraphs, this will be expressly indicated at the appropriate point. In this case, the special terms of use apply in each individual case.
Source: Imprint template from JuraForum.de and translated with DeepL.
In accordance with §28 III of the Federal Data Protection Act, we prohibit the use of our data for advertising purposes!