The Voice Conversion Challenge 2016

By Tomoki Toda, Linghui Chen, Daisuke Saito, Fernando Villavicencio, Mirjam Wester, Zhizheng Wu, Junichi Yamagishi. 2016

193 citations recorded by the source corpus.

Subfields: Artificial Intelligence

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Junichi Yamagishi
2. Xu Tan
3. Tao Qin
4. Tie-Yan Liu
5. Frank K. Soong
6. Haizhou Li
7. Hung-yi Lee
8. Berrak Sisman

Citations in the corpus

• MBNET: MOS Prediction for Synthesized Speech with Mean-Bias Network (2021, methodology): “In order to test the generalization ability of MBNet, we further utilize the MOS dataset from Voice Conversion Challenge 2016 (VCC 2016) [20] including 26028 audios from 20 systems as the test set.”
• Locally Linear Embedding for Exemplar-Based Spectral Conversion (2016, methodology): “The proposed LLE-exemplar-based SC system was evaluated using a parallel English speech corpus provided by the vcc2016 organizer [26].”
• Pretraining Techniques for Sequence-to-Sequence Voice Conversion (2020, background): “In VC, it is common to limit the size of training data to around 5 or 10 minutes [31], [32], wherein existing seq2seq VC literature, approximately 1 h of data is usually used.”
• Voice Conversion from Unaligned Corpora Using Variational Autoencoding Wasserstein Generative Adversarial Networks (2017, methodology): “The proposed VC system was evaluated on the Voice Conversion Challenge 2016 dataset [8].”
• Deepfakes generation and detection: state-of-the-art, open challenges, countermeasures, and way forward (2021, methodology): “Various VC challenges (VCC) have been held to encourage the development of VC techniques and improve the quality of converted speech [198-200].”
• CycleGAN-VC: Non-parallel Voice Conversion Using Cycle-Consistent Adversarial Networks (2018, methodology): “We evaluated our method on a non-parallel VC task using the VCC 2016 dataset [32], which was recorded by professional US English speakers, including five females and five males.”
• Analysis of the Voice Conversion Challenge 2016 Evaluation Results (2016, background): “Q indicate the 17 participants [1] and S = source, T = target and B = baseline.”
• Emotional Voice Conversion: Theory, Databases and ESD (2021, background): “For example, VCC 2016 database (Toda et al., 2016) contains parallel speech data from 5 male speakers and 5 female speakers, with each speaker speaking 216 utterances in total.”
• An Overview of Voice Conversion and Its Challenges: From Statistical Modeling to Deep Learning (2020, methodology): “conversion challenge (VCC) 2016, which was first launched [67]–[69] at INTERSPEECH 2016.”
• Statistical Parametric Speech Synthesis Incorporating Generative Adversarial Networks (2017, background): “However, the speech parameters generated from these models tend to be over-smoothed, and the resultant quality of speech is still low compared with that of natural speech [1], [9].”
• Voice conversion from non-parallel corpora using variational auto-encoder (2016, methodology): “1) The VCC2016 Speech Corpus: The proposed SC system was evaluated on a parallel English corpus from the Voice Conversion Challenge 2016 [20].”
• Deep Learning Based Assessment of Synthetic Speech Naturalness (2020, background): “The authors would like to thank the organizers of the Blizzard Challenge and the Voice Conversion Challenge for making the results of past challenges available.”
• Utilizing Self-supervised Representations for MOS Prediction (2021, methodology): “Our framework surpasses the two previous state-of-the-art models on Voice Conversion Challenge (VCC) 2018 [19] while being at least comparable on VCC 2016 [20].”
• On Using Backpropagation for Speech Texture Generation and Voice Conversion (2017, background): “An overview of the state-of-the-art in this area can be seen in the recent Voice Conversion Challenge [35].”