International Audio Laboratories Erlangen

Johannes Zeitler, M.Sc.

Research Interests

  • Audio synchronization using neural features
  • Differentiable alignment techniques

CV

  • 10/2022 – Present: PhD Student at the International Audio Laboratories Erlangen
  • 10/2021 – 10/2022: Research Assistant at Chair of Multimedia Communications and Signal Processing at FAU Erlangen
  • 04/2019 – 08/2021: M.Sc. in Advanced Signal Processing and Communications Engineering at FAU Erlangen
  • 10/2015 – 04/2019: B.Sc. in Electrical Engineering at FAU Erlangen

Publicity

Teaching

  • Assistant for seminar "Audio Processing Seminar" (Winter 22/23, Summer 23)
  • Assistant for lab course "Audio Processing Laboratory" (Winter 22/23, Summer 23, Winter 23/24)
  • Instructor for Research Internships (10 ECTS) on "DNN-based music analysis and synthesis" (Summer 2024)
  • Instructor for Research Internships (10 ECTS) on "DNN-based music analysis with Pytorch" (Summer 2025)

Thesis and Project Supervision

  • Simon Deniffel. Hybrid Models for Music Information Retrieval Tasks. Major Research Project, 2023.
  • Celia Birle. Estimating Pitch-Class Onsets from Music Recordings using Low-Complexity Neural Networks. Research Internship, 2024.
  • Purushotham Koduri. VAE-ry Ordinary Audio Synthesizer: Exploring Variational Autoencoders for Waveform Synthesis. Research Internship, 2024.
  • Quang Hoang Nguyen Vo. Training with Unaligned Datasets: Soft Dynamic Time Warping. Research Internship, 2025.
  • Yijiong Wang. Template-Based Chord Recognition: Training with Weakly Aligned Data Using the CTC Algorithm. Research Internship, 2025.
  • Zahra Zamanoghli-Taleschi. Differentiable Harmonic-Percussive Source Separation in Pytorch. Research Internship, 2025.
  • Fatemeh Hosseini. A Deep Learning Approach to Onset Detection Using Differentiable Spectral Flux. Research Internship, 2025.
  • Mika Kurz. Differentiable Template-Based F0 Estimation. Research Internship, 2025.
  • Ole Müermann. Expanding Chord Recognition Beyond 24 Classes via Deep Learning. Research Internship, 2025.
  • Peter Kodl. Computer-Assisted Visualization of Harmonic Structures in Music Recordings: A Case Study on Beethoven Piano Sonatas. Bachelor Thesis, 2025.
  • Yijiong Wang. Motif-Informed Audio Representation Learning for Piano Music Using Differentiable Dynamic Time Warping. Master Thesis, 2026.
  • Quang Hoang Nguyen Vo. Training Piano Music Transcription Models on Weakly Aligned Data Using Differentiable Dynamic Time Warping. Master Thesis, 2026.

Code

  • Beethoven Piano Sonata Dataset (BPSD): Scripts [GitHub]
  • Stabilizing Training With Soft Dynamic Time Warping [GitHub]
  • Soft Dynamic Time Warping With Variable Step Weights [GitHub]
  • Subsequence Soft Dynamic Time Warping [Github]
  • A Unified Perspective on CTC and SDTW using Differentiable DTW [GitHub]

Resources

  • Accompanying website for paper "Reformulating Soft Dynamic Time Warping: Insights Into Target Artifacts and Prediction Quality"

Datasets