Zobrazeno 1 - 10
of 305
pro vyhledávání: '"Erdogan, Hakan"'
Autor:
Erdogan, Hakan, Wisdom, Scott, Chang, Xuankai, Borsos, Zalán, Tagliasacchi, Marco, Zeghidour, Neil, Hershey, John R.
We present TokenSplit, a speech separation model that acts on discrete token sequences. The model is trained on multiple tasks simultaneously: separate and transcribe each speech source, and generate speech from text. The model operates on transcript
Externí odkaz:
http://arxiv.org/abs/2308.10415
Autor:
Yang, Yang, Shih, Shao-Fu, Erdogan, Hakan, Lin, Jamie Menjay, Lee, Chehung, Li, Yunpeng, Sung, George, Grundmann, Matthias
High quality speech capture has been widely studied for both voice communication and human computer interface reasons. To improve the capture performance, we can often find multi-microphone speech enhancement techniques deployed on various devices. M
Externí odkaz:
http://arxiv.org/abs/2303.07486
Autor:
Muckenhirn, Hannah, Safin, Aleksandr, Erdogan, Hakan, Quitry, Felix de Chaumont, Tagliasacchi, Marco, Wisdom, Scott, Hershey, John R.
Typically, neural network-based speech dereverberation models are trained on paired data, composed of a dry utterance and its corresponding reverberant utterance. The main limitation of this approach is that such models can only be trained on large a
Externí odkaz:
http://arxiv.org/abs/2203.15652
Autor:
Adiguzel, Seyfure, Karamese, Miray, Kugu, Senanur, Kacar, Elif Ayse, Esen, Muhammed Fevzi, Erdogan, Hakan, Tasoglu, Savas, Bacanli, Merve Güdül, Altuntas, Sevde
Publikováno v:
In International Journal of Biological Macromolecules October 2024 278 Part 4
The recently-proposed mixture invariant training (MixIT) is an unsupervised method for training single-channel sound separation models in the sense that it does not require ground-truth isolated reference sources. In this paper, we investigate using
Externí odkaz:
http://arxiv.org/abs/2110.10739
Autor:
Erdoğan, Hakan, Bacanlı, Merve Güdül, Karayavuz, Burcu, Eşim, Özgür, Sarper, Meral, Erdem, Onur, Özkan, Yalçın
Publikováno v:
In Journal of Drug Delivery Science and Technology August 2024 97
Autor:
Koizumi, Yuma, Karita, Shigeki, Wisdom, Scott, Erdogan, Hakan, Hershey, John R., Jones, Llion, Bacchiani, Michiel
Single-channel speech enhancement (SE) is an important task in speech processing. A widely used framework combines an analysis/synthesis filterbank with a mask prediction network, such as the Conv-TasNet architecture. In such systems, the denoising p
Externí odkaz:
http://arxiv.org/abs/2106.15813
Supervised neural network training has led to significant progress on single-channel sound separation. This approach relies on ground truth isolated sources, which precludes scaling to widely available mixture data and limits progress on open-domain
Externí odkaz:
http://arxiv.org/abs/2106.00847
Autor:
Erdoğan, Hakan, Karayavuz, Burcu, Bacanlı, Merve Güdül, Eşim, Özgür, Sarper, Meral, Altuntaş, Sevde, Erdem, Onur, Özkan, Yalçın
Publikováno v:
In Journal of Photochemistry & Photobiology, B: Biology April 2024 253
Autor:
Maiti, Soumi, Erdogan, Hakan, Wilson, Kevin, Wisdom, Scott, Watanabe, Shinji, Hershey, John R.
Publikováno v:
ICASSP 2021, SPE-54.1
We present an end-to-end deep network model that performs meeting diarization from single-channel audio recordings. End-to-end diarization models have the advantage of handling speaker overlap and enabling straightforward handling of discriminative t
Externí odkaz:
http://arxiv.org/abs/2105.02096