CV
| Experience | ||
| Audio Technologies Researcher at Huawei Technologies, Munich Research Center | ||
| 01.2017 – present | Principal Research Engineer | |
| Audio technology specialist for media playback, speech enhancement, AR/VR and Audio Intelligence applications for smartphones Designing DSP solutions for loudspeaker correction and protection, microphone array speech processing, and VR 3D audio rendering Developing AI systems for media content and music genre recognition (DNN), acoustic scene and environment recognition (SVM) and source separation (LSTM) Ensuring deliverables from R&D of next generation audio solutions to products Implementation of algorithms in Matlab, Pyton, float C utilizing TensorFlow, Keras, CURRENNT, and TF Lite toolkits |
||
| 01.2015 – 12.2016 | Senior Research Engineer | |
| Developing DSP solutions for sound playback enhancement (micro-speaker and headphones) and voice communication (uplink and downlink) Extending SWS: incorporating algorithms for stereo to 5 channel upmixing, dialogue enhancement, and crosstalk cancellation Researching speech enhancement algorithms for voice communication, microphone array noise reduction and speech intelligibility enhancement Implementing algorithms in Matlab and float C; developing real-time demonstrators using VST and Android SDK/NDK toolkits Delivering algorithms to tablets (Mediapads M2, M3) and smartphones (Mate 9) Managing technical collaboration projects with universities and SMEs Securing IPR, filing 30+ PCT applications |
||
| 11.2012 – 12.2014 | Research Engineer 3D Audio | |
| Designing a complete DSP solution for sound playback enhancement with increased sound quality on smartphones Delivering the Huawei SWS (Super Wide Sound) sound enhancement solution to Huawei Mediapad products (replacing the omnipresent DTS suite) Developing algorithms: EQ, DRC, limiting, bass enhancement (missing fundamental), binaural rendering and crosstalk cancellation Rapid prototyping, product related research and demonstration of audio DSP algorithms to stakeholders and product lines Implementing a complete sound enhancement solution satisfying product requirements (Matlab), writing efficient code meeting hardware limitations (floating point C) and assisting fixed-point integration (Tensilica HiFi) |
||
| Music Information Retrieval Researcher at Saarland University, Saarbrücken | ||
| 06.2008 – 10.2012 | Research and teaching in the field of audio and music signal processing and music information retrieval | |
| Developing audio DSP algorithms for music analysis, feature extraction, and efficient content-based search in large music databases | ||
| Researching beat tracking and tempo estimation algorithms which can cope with time variying tempo and classical music without drums; repetition based music segmentation and structure analysis (chorus detection); music retrieval and cover song recognition based on Locality-sensitive hashing (LSH) for efficient similarity search; music transcription and chord recognition | ||
| Research intern at NYU Music and Audio Research Laboratory, New York | ||
| 10.2010 – 11.2010 | Developed an unsupervised musical motif discovery system using sparse convolutive Non-Negative Matrix Factorization (NMF) in Python | |
| Research intern at the Audio & Multimedia Department at Fraunhofer IIS, Erlangen | ||
| 02.2007 – 05.2007 | Surround Upmix: Development of algorithms for converting stereo to 5.1 (Matlab) | |
| Instrument recognition in polyphonic music signals based on Hidden Markov Models (HMM) using HTK Speech Recognition toolkit | ||
| Research assistant at the Institute for Human-Machine Communication, TUM | ||
| 12.2005 – 01.2006 | Assisted in building a multimodal database for developing and evaluating emotion recognition procedures |
|
| Intern at Akustikbüro Schwartzenberger & Burkhart, Pöcking, Germany. | ||
| 02.2005 – 04.2005 | Room acoustic measurements, simulations and optimizations | |
| Education | ||
| PhD in Computer Science at Saarland University and Max-Planck-Institut Informatik, Saarbrücken, Germany | ||
| 06.2008 – 11.2012 | Title of PhD thesis: "Signal Processing Methods for Beat Tracking, Music Segmentation, and Audio Retrieval" | |
| 20+ publications at scientific conferences, journals and books | ||
| Supervised by Prof. Meinard Müller | ||
| Degree: Dr.-Ing, Grade: magna cum laude (very good) | ||
| Master in Electrical Engineering and Information Technology at TU München, Germany | ||
| 10.2004 – 05.2008 | Main focus on Digital Signal Processing, Pattern Recognition, Information Retrieval and Human-Machine Interaction; Personal focus on audio and music analysis | |
| Diploma thesis: "Polyphonic Music Transcription based on Real Audio" Supervised by Prof. Björn Schuller Developed a complete system deriving a symbolic MIDI representation from a polyphonic audio recording; combining mulit-pitch estimation, onset detection and Hidden Markov Models (HMM) for pitch tracking. |
||
| Degree: Dipl.-Ing. Univ., Grade: 1.5 (very good) | ||
| Studies abroad: Lunds Universitet, Lund, Sweden | ||
| 08.2006 – 01.2007 | Selected courses in Machine Learning and statistics | |
| Undergraduate studies in Electrical Engineering and Information Technology at TU München, Germany | ||
| 10.2002 – 10.2004 | Main focus on Information Technology and Communication Engineering | |
| Bachelor thesis: "An Application for Automatically Organizing, Structuring, and Analyzing Large Music Collections" | ||
| Developed a foobar2000 component adding genre recognition and music similarity search functionality based on custom feature extraction and SVM classification; Tools: Weka, LIBSVM, C++ | ||
| Degree: B.Sc., Grade: 1.6 (good) | ||