Zobrazeno 1 - 10
of 1 087
pro vyhledávání: '"P. Vaeßen"'
Autor:
Vaessen, Nik, van Leeuwen, David A.
Foundation models in speech are often trained using many GPUs, which implicitly leads to large effective batch sizes. In this paper we study the effect of batch size on pre-training, both in terms of statistics that can be monitored during training,
Externí odkaz:
http://arxiv.org/abs/2402.13723
Autor:
Vaessen, T., van Roestel, J.
Context. Double eclipsing binaries are gravitationally bound quadruple systems in a 2+2 configuration where both of the binaries are eclipsing. These systems are interesting objects to better understand stellar formation, to investigate the dynamical
Externí odkaz:
http://arxiv.org/abs/2312.08529
Publikováno v:
2023 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2023)
Recent research has proposed approaches that modify speech to defend against gender inference attacks. The goal of these protection algorithms is to control the availability of information about a speaker's gender, a privacy-sensitive attribute. Curr
Externí odkaz:
http://arxiv.org/abs/2306.17700
As mobile health (mHealth) studies become increasingly productive due to the advancements in wearable and mobile sensor technology, our ability to monitor and model human behavior will be constrained by participant receptivity. The reliance on subjec
Externí odkaz:
http://arxiv.org/abs/2303.09077
Autor:
Vaessen, Nik, van Leeuwen, David A.
We study multi-task learning for two orthogonal speech technology tasks: speech and speaker recognition. We use wav2vec2 as a base architecture with two task-specific output heads. We experiment with different architectural decisions to mix speaker a
Externí odkaz:
http://arxiv.org/abs/2302.12773
We investigate recent transformer networks pre-trained for automatic speech recognition for their ability to detect speaker and language changes in speech. We do this by simply adding speaker (change) or language targets to the labels. For Wav2vec2 p
Externí odkaz:
http://arxiv.org/abs/2302.09381
Publikováno v:
Stress, Vol 27, Iss 1 (2024)
Laboratory stress tasks are necessary to closely investigate the stress response in a controlled environment. However, to our knowledge, no study has tested whether participating in such tasks can pose any daily life adverse effect. Fifty-three healt
Externí odkaz:
https://doaj.org/article/c9ce32b093d34ab7877245e2d3e7e5ee
Publikováno v:
International Journal of Electrical Power & Energy Systems, Vol 158, Iss , Pp 109911- (2024)
The correct identification of partial discharges (PDs) is instrumental for the maintenance plan in gas-insulated systems (GIS). However, onsite PD measurements are difficult, especially in HVDC systems, where partial discharges can be confused with i
Externí odkaz:
https://doaj.org/article/a49e8b2cdfc049f3a401f40885776691
On-site soil analysis: A novel approach combining NIR spectroscopy, remote sensing and deep learning
Publikováno v:
Geoderma, Vol 446, Iss , Pp 116903- (2024)
Soil health is essential to global sustainable food production. Beyond its role in food production, soil also plays a crucial role in maintaining ecosystem health and mitigating climate change. Monitoring and improving the health of agricultural soil
Externí odkaz:
https://doaj.org/article/68052fbfb05c4e0981112436f16c1521
Autor:
Vaessen, Nik, van Leeuwen, David A.
Publikováno v:
Proc. Interspeech 2022, 4760-4764
This work considers training neural networks for speaker recognition with a much smaller dataset size compared to contemporary work. We artificially restrict the amount of data by proposing three subsets of the popular VoxCeleb2 dataset. These subset
Externí odkaz:
http://arxiv.org/abs/2203.14688