no code implementations • 19 Feb 2024 • Haolin Chen, Philip N. Garner
Our results demonstrate that catastrophic forgetting can be overcome by our methods without degrading the fine-tuning performance, and using the Kronecker factored approximations produces a better preservation of the pre-training knowledge than the diagonal ones.
no code implementations • 29 Nov 2023 • Pavel Korshunov, Haolin Chen, Philip N. Garner, Sebastien Marcel
From the publicly available speech dataset LibriTTS, we also created a separate database of only audio deepfakes LibriTTS-DF using several latest text to speech methods: YourTTS, Adaspeech, and TorToiSe.
no code implementations • 3 Mar 2023 • Haolin Chen, Philip N. Garner
Given the recent success of diffusion in producing natural-sounding synthetic speech, we investigate how diffusion can be used in speaker adaptive TTS.
3 code implementations • 7 Mar 2022 • Florian Mai, Arnaud Pannatier, Fabio Fehr, Haolin Chen, Francois Marelli, Francois Fleuret, James Henderson
We find that existing architectures such as MLPMixer, which achieves token mixing through a static MLP applied to each feature independently, are too detached from the inductive biases required for natural language understanding.
no code implementations • 4 Nov 2021 • XiaoHui Yang, Zheng Wang, Huan Wu, Licheng Jiao, Yiming Xu, Haolin Chen
The proposed model aims to mine the hidden semantic information and intrinsic structure information of all available data, which is suitable for few labeled samples and proportion imbalance between labeled samples and unlabeled samples problems in frontal face recognition.
no code implementations • 4 Nov 2020 • Ying Shi, Haolin Chen, Zhiyuan Tang, Lantian Li, Dong Wang, Jiqing Han
Recently, speech enhancement (SE) based on deep speech prior has attracted much attention, such as the variational auto-encoder with non-negative matrix factorization (VAE-NMF) architecture.
no code implementations • 16 Jul 2020 • Haolin Chen, Luis Rademacher
We propose a new algorithm for tensor decomposition, based on Jennrich's algorithm, and apply our new algorithmic ideas to blind deconvolution and Gaussian mixture models.
2 code implementations • 31 Oct 2019 • Yue Fan, Jiawen Kang, Lantian Li, Kaicheng Li, Haolin Chen, Sitong Cheng, Pengyuan Zhang, Ziya Zhou, Yunqi Cai, Dong Wang
These datasets tend to deliver over optimistic performance and do not meet the request of research on speaker recognition in unconstrained conditions.