CheapTrick, a spectral envelope estimator for high-quality speech synthesis
By Masanori Morise. 2015
109 citations recorded by the source corpus.
Abstract A spectral envelope estimation algorithm is presented to achieve high-quality speech synthesis. The concept of the algorithm is to obtain an accurate and temporally stable spectral envelope. The algorithm uses fundamental frequency (F0) and consists of F0-adaptive windowing, smoothing of the power spectrum, and spectral recovery in the quefrency domain. Objective and subjective evaluations were carried out to demonstrate the effectiveness of the proposed algorithm. Results of both evaluations indicated that the proposed algorithm can obtain a temporally stable spectral envelope and synthesize speech with higher sound quality than speech synthesized with other algorithms.
Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Signal Processing
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
- Junichi Yamagishi
- Xin Eric Wang
- Tomoki Toda
- Shinji Takaki
- Yi-Chiao Wu
- Gustav Eje Henter
- Zhenhua Ling
- Kenji Ozawa
In Computer Vision and Pattern Recognition
- Junichi Yamagishi
- Shinji Takaki
- Kenji Ozawa
- Fumiya Yokomori
- Xin Eric Wang
- Gustav Eje Henter
- Tomoki Toda
- Reo Yoneyama
In Signal Processing
- Junichi Yamagishi
- Xin Eric Wang
- Tomoki Toda
- Shinji Takaki
- Yi-Chiao Wu
- Gustav Eje Henter
- Kenji Ozawa
- Fumiya Yokomori
Citations in the corpus
- WORLD: A Vocoder-Based High-Quality Speech Synthesis System for Real-Time Applications (2016, methodology): “This is in contrast to past evaluations of conventional systems that used speech consisting of only vowels [16], [17].”
- Unified Source-Filter GAN: Unified Source-filter Network Based On Factorization of Quasi-Periodic Parallel WaveGAN (2021, methodology): “We use a simplified algorithm of cheaptrick [33] to extract the spectral envelope from the output signal of source-network.”
- A deep auto-encoder based low-dimensional feature extraction from FFT spectral envelopes for statistical parametric speech synthesis (2016, methodology): “For instance, a spectral envelope estimation algorithm implemented in the WORLD vocoder performs F0-adaptive windowing, smoothing of the power spectrum, and spectral recovery in the quefrency domain for obtaining accurate and stable spectral envelopes [13, 14].”
- Error Evaluation of an F0-Adaptive Spectral Envelope Estimator in Robustness against the Additive Noise and F0 Error (2015, methodology): “The results of subjective evaluation demonstrated that CheapTrick was superior to the conventional algorithms in terms of the sound quality of synthesized speech and robustness of the temporal F0 change [17].”
- Deep Encoder-Decoder Models for Unsupervised Learning of Controllable Speech Synthesis (2018, methodology): “In particular, Open JTalk [86] was used to extract 389 linguistic features while WORLD [75], [87] was used for acoustic analysis and signal synthesis.”