βοΈ l.sanner@gmx.de π linked.in/lukas-sanner π¨βπ»github.com/Farbdrucker |
Open minded, enthusiastic about continuous learning and data. Working autonomously and always looking for new challenges.
I always wanted to have some smart assistant but it is way more fun to create one yourself but buying a Alexa or what so ever. So, here is my attempthttps://t.co/woznaRzEAQ pic.twitter.com/sojlBb3RkY
β Sanner (@scanner_c) July 12, 2020
Had fun last weekend at #diffusion2019 with @cybertreiber experimenting with encrypted machine learning #crypten and actually contributing to it. Also won a small price from @oceanprotocol.https://t.co/0DeQ9AMZAV
β Sanner (@scanner_c) October 26, 2019
Everyone needs faster annotation for #ComputerVision tasks, that's why we have a deeply learned magic wand in the making! Read about the first version here by @LukasSanner https://t.co/qUNudeaiUI
β Alexander Hirner (@cybertreiber) April 10, 2019
4th place at ZEISShack2019 with a photomask defect detection method
Platz 2 beim #automationhackathon 2018 @spsipcdrives - unser Team vom #StartUp MoonVision aus Wien.
β Phoenix Contact DE (@PhoenixContactV) November 27, 2018
Aber was genau haben sie eigentlich gehackt? Die Antwort spΓ€ter π€π #PLCnext #sps_live pic.twitter.com/yq4wltslkR
Implemented a tool to reproduce and compare DL trainings easily with
different hyper parameters, introduced an active learning workflow by
prediction uncertainty (with pending patent application), successfully
managed projects with leading customers in the industry, supervisor for
interns. (Python
: opencv
,
pytorch
, numpy
, prefect
)
06.2020 - present | Head of Computer Vision
Solutions Designing and implementing a scalable infrastructure for DL training and evaluation and customer project lead |
10.2018 - 05.2020 | Research Engineer Researching, implementing and testing newest CV papers and algorithms |
Developed a tool to enable automatic equalizer settings for podcasts using deep learning
11.2017 - 03.2018 | Intern - Computer
Scientist Research for DL techniques to improve sound, thereby developed own CNN model with Tensorflow, C++ implementations of signal processing algorithms ( C++ : boost ,
Intel MKL , Python : Librosa ,
TensorFlow , Caffe ) |
Focus on numerical linear Algebra and ordinary differential
equations, especially eigenvalue algorithms on large (sparse/dense)
symmetric positive semidefinite matrices. (Matlab
,
Python
: FEniCS
, Numpy
)
04.2016 - 08.2018 | Maser of Science, Applied
Mathematics Thesis Comparison of Eigenvalue Algorithms for the Graph Laplacian with Application to Semi-Supervised Learning |
10.2011 - 04.2016 | Bachelor of Science,
Mathematics Thesis Vorkonditionierung der instationaeren Navier-Stokes Gleichung |
π¨βπ» Coding Project Euler (Rust
,
C++
), refactoring of bachelor thesis (Python
),
rewriting master thesis (C++
), Maker: LED micro controller
(C
)
πΊ Music learning contrabass, playing trumpet, electric-guitar and -bass
π§ Sports bouldering, running
πΈ Interests photography, hiking, espresso
Made with pandoc
Β© Lukas Sanner, 2020