no code implementations • 6 Sep 2024 • Mattes Ohlenbusch, Christian Rollwage, Simon Doclo
Hearable devices, equipped with one or more microphones, are commonly used for speech communication.
no code implementations • 19 May 2024 • Mattes Ohlenbusch, Christian Rollwage, Simon Doclo
The proposed techniques use few recorded own voice signals to estimate transfer characteristics and can then be used to simulate a large amount of own voice signals based on single-channel speech signals.
no code implementations • 14 Dec 2023 • Mattes Ohlenbusch, Christian Rollwage, Simon Doclo
Recording a sufficient amount of noise required for training such a system is costly since noise transmission between outer and inner microphones varies individually.
no code implementations • 10 Oct 2023 • Mattes Ohlenbusch, Christian Rollwage, Simon Doclo
In this paper, we propose a speech-dependent model of the own voice transfer characteristics based on phoneme recognition, assuming a linear time-invariant relative transfer function for each phoneme.
no code implementations • 15 Sep 2023 • Mattes Ohlenbusch, Christian Rollwage, Simon Doclo
To enhance the quality of the in-ear microphone signal using algorithms aiming at joint bandwidth extension, equalization, and noise reduction, it is desirable to have an accurate model of the own voice transfer characteristics between the entrance of the ear canal and the in-ear microphone.
no code implementations • 19 Apr 2023 • Paul M. Reuter, Christian Rollwage, Bernd T. Meyer
Our system achieves a promising accuracy for streaming keyword spotting and keyword search on Common Voice audio using just 5 examples per keyword.
no code implementations • 27 May 2022 • Ragini Sinha, Marvin Tammen, Christian Rollwage, Simon Doclo
Target speaker extraction aims at extracting the target speaker from a mixture of multiple speakers exploiting auxiliary information about the target speaker.
no code implementations • 12 May 2022 • Mattes Ohlenbusch, Christian Rollwage, Simon Doclo
In this paper, we apply a deep learning-based bandwidth-extension system to the own voice reconstruction task and investigate different training strategies in order to overcome the limited availability of training data.
no code implementations • 9 Apr 2021 • Ragini Sinha, Marvin Tammen, Christian Rollwage, Simon Doclo
In this paper, we focus on a single-channel target speaker extraction system based on a CNN-LSTM separator network and a speaker embedder network requiring reference speech of the target speaker.