Pith. sign in

REVIEW 4 major objections 4 minor 38 references

CSSinger: End-to-End Chunkwise Streaming Singing Voice Synthesis System Based on Conditional Variational Autoencoder

T0 review · 4 major / 4 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A chunkwise streaming singing-voice synthesizer built on a conditional variational autoencoder matches or outperforms a fully parallel baseline in quality and pitch accuracy, while reducing first-audio latency from 0.18 to 0.05 seconds.

desk verdict Genuinely useful streaming SVS system with a clever padding fix, but the evaluation numbers are inconsistent and the significance claims overreach. read the letter →

arxiv 2412.08918 v2 pith:6EN5FPCM submitted 2024-12-12 eess.AS

Jianwei Cui , Yu Gu , Shihao Chen , Jie Zhang , Liping Chen , Lirong Dai This is my paper · ORCID
classification eess.AS
keywords singingvoicesynthesisstreamingconditionalvariationalautoencoderchunkwisegenerationmemory-bankattentioncausalconvolutionnaturalpaddinglow-latencytext-to-speech
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to show that singing voice synthesis—normally computed over an entire song at once—can be made to run chunk by chunk without losing the quality or pitch accuracy of a fully parallel system. The authors build a conditional variational autoencoder that generates audio through latent representations, then make both the acoustic decoder and the neural vocoder stream in fixed-size chunks. Their fully streaming system, CSSinger-FS, scores higher than the parallel SiFiSinger baseline in subjective listening tests on two singing datasets, and matches or beats it on objective pitch metrics, while cutting first-audio latency from 0.18 seconds to 0.05 seconds on GPU. The same recipe transfers to text-to-speech. The practical point is that real-time singing synthesis on phones or cloud endpoints may not require a quality sacrifice.

What carries the argument

The load-bearing mechanism is the ChunkStream Decoder, a chunked attention decoder that processes each fixed-length chunk of the latent sequence independently, drawing key and value context from a small left window, a small right window, and a memory bank carried over from earlier chunks; the tail of each chunk is passed forward through a Causal Smooth Layer built from two 1-D causal convolutions and layer normalization, so attention is quadratic only inside a chunk while information still flows across chunk boundaries. The second mechanism is Natural Padding in the causal vocoder generator: the left-side padding on causal convolutions and transposed convolutions is filled with actual latent values from preceding frames rather than zeros, and the output is trimmed from the tail to the expected upsampled length. The paper also replaces the posterior encoder's convolutions with causal versions. These components work together to remove the training/inference mismatch that appears when a latent code with no fixed semantic meaning is sliced into chunks and fed to a causal streaming network.

What would settle it

Take a song with a fast pitch leap deliberately placed exactly on a chunk boundary and one placed in the middle of a chunk: CSSinger-FS should show no extra F0 error on the boundary version if the Causal Smooth Layer does its job, and a large boundary-specific error spike would falsify the paper's central claim.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that latent representations from a conditional VAE can be streamed through a causal vocoder if two mismatches are fixed. First, the causal generator must be padded with real latent values drawn from preceding frames, what the paper calls 'natural padding,' instead of the constant padding (zeros) that works for mel-spectrograms but corrupts learned latent codes. Second, the acoustic model decoder must itself be chunkwise, using memory-bank attention plus a Causal Smooth Layer, so that the prior distribution is produced in the same chunk-wise pattern the generator consumes. With these fixes, the fully streaming model achieves the best mean opinion score on Opencpop and PopCS among all compared systems (3.607 and 3.500), lowers F0 RMSE from 34.295 to 28.601 on Opencpop relative to the parallel baseline, and is the only system in the paper's example that renders a high upward pitch transition at a phrase ending. The paper's claim is that chunkwise streaming need not cost expressiveness: aligning the decoder's chunks with the vocoder's causal consumption of latents can slightly improve pitch accuracy while drastically lowering latency.

Load-bearing premise

The chunkwise decoder only sees a small local context—twenty frames per chunk, ten frames of left context, and four frames of right context—plus a memory bank, so the whole method stands on the assumption that this is enough musical context to keep pitch and prosody smooth across chunk boundaries.

Editorial extensions

If this is right

  • First-audio latency drops from 0.180 s (parallel baseline) to 0.051 s on GPU and from 1.508 s to 0.483 s on CPU, so the streaming system can serve real-time, batch-transmitted online synthesis.
  • On the Opencpop and PopCS singing datasets, CSSinger-FS improves mean opinion score over the fully parallel SiFiSinger baseline (e.g., 3.607 vs 3.510 on Opencpop), indicating chunkwise generation does not cost subjective quality.
  • Pitch accuracy improves as well: F0 RMSE falls from 34.295 to 28.601 on Opencpop, and the only system that renders a final-note upward pitch transition in the paper's example is the fully streaming model.
  • The same fully streaming design transfers to TTS: on the Baker dataset it reaches MOS 3.828, just below the parallel baseline's 3.911 while streaming.
  • The ablation attributes a large F0 RMSE penalty (37.872 vs 28.101) to removing the Causal Smooth Layer, so boundary smoothing is essential to the pitch result.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • As an editorial extension, Natural Padding should transfer to any neural codec or latent-conditioned effect where causal convolution touches a learned latent space; the fix is generic and not limited to singing.
  • The fixed chunk and context sizes (20 frames per chunk, 10 frames left, 4 frames right) imply a tunable latency-quality knob; larger chunks should narrow the remaining gap to the parallel baseline, so real deployments could trade a few tens of milliseconds for more context.
  • The paper leaves real-time input prediction untested; if the front-end also streams pitch, duration, and lyric alignment, the full system could close an actual live-feedback loop, but that remains an extension, not a claim here.
  • Because the causal posterior encoder ablation mainly affects high-note transitions, the prior and posterior causalization likely regularizes the learned latent trajectory itself; that suggests streaming quality may depend as much on what the model is trained to encode as on the decoder's context window.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. The paper proposes CSSinger, an end-to-end chunkwise streaming singing voice synthesis system built on the SiFiSinger conditional VAE. The contributions are threefold: a causal HiFi-GAN generator with a 'natural padding' technique for latent features, a ChunkStream Decoder that generates latent representations chunk-by-chunk using Emformer-style attention with a causal smooth layer, and a fully streaming inference pipeline. Experiments on Opencpop, PopCS, and Baker report MOS and objective metrics for SVS and TTS, latency measurements on GPU/CPU configurations, ablations, and a horizontal comparison with VISinger2 and DiffSinger. The central claim is that CSSinger-FS matches or beats the parallel SiFiSinger baseline while reducing latency.

Significance. If the results are reproducible, the work is a meaningful step toward practical low-latency end-to-end SVS: it identifies a real train/test mismatch when causal vocoders consume learned latent features and offers a simple fix (natural padding); it adapts chunkwise attention to latent generation; and it provides latency benchmarks across hardware configurations. The architecture is plausible, the demo allows auditory verification, and the ablation study gives useful evidence for the causal smooth layer and natural padding. However, the central comparative claims are currently undermined by inconsistent reported numbers for the same model and by significance statements that the confidence intervals do not support. The strengths of the paper are the engineering design and the latency analysis; the weaknesses are in the evidence quality and the precision of the claims.

major comments (4)
  1. [Main Results & Analysis, Tables 1/2/7/8] The same model, CSSinger-FS, receives inconsistent scores across tables for the same dataset. On Opencpop, Table 1 reports MOS 3.607±0.091 while Table 8 reports MOS 4.07±0.081; Table 2 reports F0 RMSE 28.601 while Table 7 reports 28.101. The text states that objective metrics are computed on all test-set entries, so the F0 discrepancy cannot be due to sample selection. Subjective MOS can differ across listener groups, but the paper does not state that Table 8 uses different listeners or samples. Please reconcile the evaluation protocols, checkpoints, or report consistent numbers; as written, the reader cannot tell which set of numbers describes the proposed model.
  2. [Evaluation of SVS, Table 1; Evaluation on TTS, Table 5] The claim that CSSinger-FS 'significantly outperforms' the parallel baseline is not supported by the reported 95% confidence intervals. On PopCS, CSSinger-FS (3.500±0.086) and SiFiSinger (3.436±0.081) overlap; on Opencpop they also overlap (3.607±0.091 vs. 3.510±0.097); on Baker, CSSinger-FS is lower (3.828±0.071 vs. 3.911±0.088). To support 'significantly outperforms,' please provide a paired significance test or bootstrap confidence intervals on per-sample scores, or revise the wording to 'comparable to.'
  3. [Evaluation on Text-to-Speech, Table 6] On the Baker TTS task, CSSinger-FS is worse than SiFiSinger on every objective metric: F0 RMSE 41.278 vs. 38.187, F0 Corr 0.783 vs. 0.817, U/UV Err 0.205 vs. 0.144, MSE 1.427 vs. 1.255, and MCD 8.763 vs. 7.706. This contradicts the abstract's claim of high pitch accuracy in TTS and the introduction's statement that the proposed model 'surpass[es] or [is] on par with' baselines on one TTS dataset. Please either provide evidence supporting the TTS claim or restrict the claim to SVS and to comparisons among CSSinger variants.
  4. [Natural Padding, Figure 2] The natural padding procedure is specified only for training: the input slice of z is extended with additional z values. The inference-time procedure for the first chunk is not described. In a streaming setting, no previous latent is available to supply natural padding for the initial chunk; if zeros or another constant are used there, the train/test mismatch that the authors identify for constant padding reappears for the first chunk. Please specify the first-chunk padding mechanism and any warm-up strategy, and report whether it affects the quality or latency of the first output chunk.
minor comments (4)
  1. [Implementation Details] The sentence 'This also determines the length of the Key and Value vectors (K^n_{L,i}, K^n_{R,i}, V^n_{L,i}, V^n_{R,i}) from the left and right contexts (e.g., see Section )' contains a missing section reference; please complete it.
  2. [Main Results & Analysis, Table 1 discussion] In the paragraph after Table 1, the text says 'the introduction of Natural Padding (CSSinger-SS) can partially overcome' the degradation; based on the table and the earlier method names, this should be CSSinger-SS-NP.
  3. [Chunkwise Fully-Streaming Framework, Eqs. (5)-(12)] The memory update in Eq. (10) defines m^{n+1}_i, but the initial memory M^n_i and how it is 'obtained from the lower layer of the previous chunk' are not defined precisely; please state the dimensionality and the projection used.
  4. [Horizontal Comparison, Table 8] Please specify whether the MOS in Table 8 uses the same 20 test samples and the same listener pool as Table 1; the large difference in CSSinger-FS scores between the two tables makes this information necessary for interpretation.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the streaming components are tested against external datasets and baselines, and the self-citation to SiFiSinger is architectural lineage rather than evidence that substitutes for the paper's own measurements.

full rationale

The paper's derivation chain is self-contained. CSSinger is built on the authors' prior SiFiSinger, but the new claims (chunkwise streaming, Natural Padding, ChunkStream Decoder, Causal Smooth Layer) are evaluated by MOS and objective metrics on three external datasets (Opencpop, PopCS, Baker) against parallel and semi-streaming baselines, including SiFiSinger itself as a baseline. Natural Padding uses extra samples of the latent z as causal padding for the vocoder; the padding is the conditioning input, not the predicted audio or a fitted parameter renamed as a prediction, and the training/inference asymmetry it addresses is an engineering artifact, not a definitional equivalence. The ChunkStream Decoder explicitly borrows its memory-and-context attention from Emformer (Shi et al. 2021), an external citation, rather than importing a self-citation as a forced choice. No uniqueness theorem or ansatz is smuggled in via the authors' own prior work. The manuscript does contain reported metric inconsistencies across tables (e.g., CSSinger-FS F0 RMSE 28.601 in Table 2 vs 28.101 in Table 7, MOS 3.607±0.091 in Table 1 vs 4.07±0.081 in Table 8) and overlapping confidence intervals supporting some 'significant outperforms' phrasing; these are evidence-quality and reproducibility concerns, not circularity. No step in the paper's equations or comparisons is equivalent to its inputs by construction, so the circularity score is 0.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central claims rest on standard VAE and ELBO machinery plus several hand-chosen chunking hyperparameters and the empirical assumption that local context suffices for musical dependencies. No new physical or conceptual entities are introduced.

free parameters (5)
  • chunk_size = 20
    Chunk length for training and inference in both the ChunkStream Decoder and causal HiFi-GAN; chosen by the authors without a reported search.
  • left_context_window = 10
    Number of frames from the previous chunk used as left context in the ChunkStream Decoder attention.
  • right_context_window = 4
    Number of frames from the current chunk used as right context in the attention.
  • natural_padding_length = not precisely specified
    Padding with real z values; length is selected so that the generator output exceeds the original length, a heuristic rather than a closed-form value.
  • slice_size_for_generator_training = 20
    Random slice length for training the HiFi-GAN generator, fixed for all experiments.
assumptions (4)
  • standard math The conditional variational lower bound is a valid training objective and the posterior network provides a usable approximation.
    The paper relies on the standard ELBO derivation for conditional VAEs; this is background knowledge from VITS and SiFiSinger.
  • domain assumption Chunked attention with a memory bank and limited left/right context preserves sufficient global information for singing synthesis.
    The ChunkStream Decoder assumes that a memory bank from the previous chunk plus local context is enough to model long-range pitch and prosody dependencies.
  • domain assumption Replacing the convolutions in the posterior encoder with causal versions does not significantly affect the posterior distribution.
    The paper states this assumption without proof; it is tested only via the -Causal PostEnc ablation.
  • domain assumption MOS from 20 native speakers on 20 randomly selected samples is a reliable quality measure.
    The subjective evaluations use a small rater and sample pool without significance testing; the paper treats the resulting MOS as reliable.

how reviews work

0 comments
Cite this review

Pith. "Pith review of CSSinger: End-to-End Chunkwise Streaming Singing Voice Synthesis System Based on Conditional Variational Autoencoder." pith.science (2026). https://pith.science/paper/6EN5FPCM

@misc{pith2026241208918,
  author       = {Pith},
  title        = {Pith review of: CSSinger: End-to-End Chunkwise Streaming Singing Voice Synthesis System Based on Conditional Variational Autoencoder},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6EN5FPCM}},
  note         = {Machine review of arXiv:2412.08918}
}
read the original abstract

Singing Voice Synthesis (SVS) aims to generate singing voices of high fidelity and expressiveness. Conventional SVS systems usually utilize an acoustic model to transform a music score into acoustic features, followed by a vocoder to reconstruct the singing voice. It was recently shown that end-to-end modeling is effective in the fields of SVS and Text to Speech (TTS). In this work, we thus present a fully end-to-end SVS method together with a chunkwise streaming inference to address the latency issue for practical usages. Note that this is the first attempt to fully implement end-to-end streaming audio synthesis using latent representations in VAE. We have made specific improvements to enhance the performance of streaming SVS using latent representations. Experimental results demonstrate that the proposed method achieves synthesized audio with high expressiveness and pitch accuracy in both streaming SVS and TTS tasks.

Figures

Figures reproduced from arXiv: 2412.08918 by the authors.

Figure 1
Figure 1. Visualization of the CSSinger Model Structure: (a) Overall Training Process, (b) Semi-Streaming inference pipeline, [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Illustration of the Natural Padding Process [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. The proposed ChunkStream Decoder, where Cur [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Mel-spectrograms with phoneme boundaries and pitch contour for the same audio sample obtained by comparison [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Bar chart of CMOS scores with 95% confidence [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

38 extracted references · 22 canonical work pages

  1. [1]

    Bonada, J.; Loscos, A.; and Kenmochi, H. 2003. Sample-based singing voice synthesizer by spectral concatenation. In Proceedings of Stockholm Music Acoustics Conference, 1--4

    work page 2003

  2. [2]

    Cui, J.; Gu, Y.; Weng, C.; Zhang, J.; Chen, L.; and Dai, L. 2024. Sifisinger: A High-Fidelity End-to-End Singing Voice Synthesizer Based on Source-Filter Model. In ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 11126--11130. IEEE

    work page 2024

  3. [3]

    Dutoit, T.; and Gosselin, B. 1996. On the use of a hybrid harmonic/stochastic model for TTS synthesis-by-concatenation. Speech Communication, 19(2): 119--143

    work page 1996

  4. [4]

    Goodfellow, I.; Pouget-Abadie, J.; Mirza, M.; Xu, B.; Warde-Farley, D.; Ozair, S.; Courville, A.; and Bengio, Y. 2020. Generative adversarial networks. Communications of the ACM, 63(11): 139--144

    2020

  5. [5]

    Gu, Y.; Yin, X.; Rao, Y.; Wan, Y.; Tang, B.; Zhang, Y.; Chen, J.; Wang, Y.; and Ma, Z. 2021. ByteSing: A Chinese Singing Voice Synthesis System Using Duration Allocated Encoder-Decoder Acoustic Models and WaveRNN Vocoders. In 2021 12th International Symposium on Chinese Spoken Language Processing (ISCSLP), 1--5

    work page 2021

  6. [6]

    Hochreiter, S.; and Schmidhuber, J. 1997. Long short-term memory. Neural computation, 9(8): 1735--1780

    1997

  7. [7]

    Hono, Y.; Hashimoto, K.; Oura, K.; Nankaku, Y.; and Tokuda, K. 2019. Singing voice synthesis based on generative adversarial networks. In ICASSP 2019-2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 6955--6959. IEEE

    work page 2019

  8. [8]

    Hono, Y.; Hashimoto, K.; Oura, K.; Nankaku, Y.; and Tokuda, K. 2021. Sinsy: A deep neural network-based singing voice synthesis system. IEEE/ACM Transactions on Audio, Speech, and Language Processing, 29: 2803--2815

    work page 2021

Show all 38 references
  1. [9]

    Kalchbrenner, N.; Elsen, E.; Simonyan, K.; Noury, S.; Casagrande, N.; Lockhart, E.; Stimberg, F.; Oord, A.; Dieleman, S.; and Kavukcuoglu, K. 2018. Efficient neural audio synthesis. In International Conference on Machine Learning, 2410--2419. PMLR

    2018

  2. [10]

    Kim, J.; Choi, H.; Park, J.; Kim, S.; Kim, J.; and Hahn, M. 2018. Korean singing voice synthesis system based on an LSTM recurrent neural network. In Proc. Interspeech, 1551--1555

    2018

  3. [11]

    Kim, J.; Kong, J.; and Son, J. 2021. Conditional variational autoencoder with adversarial learning for end-to-end text-to-speech. In International Conference on Machine Learning, 5530--5540. PMLR

    2021

  4. [12]

    P.; Salimans, T.; Jozefowicz, R.; Chen, X.; Sutskever, I.; and Welling, M

    Kingma, D. P.; Salimans, T.; Jozefowicz, R.; Chen, X.; Sutskever, I.; and Welling, M. 2016. Improved variational inference with inverse autoregressive flow. Advances in neural information processing systems, 29

    2016

  5. [13]

    P.; and Welling, M

    Kingma, D. P.; and Welling, M. 2014. Auto- E ncoding V ariational B ayes. In International Conference on Learning Representations

    2014

  6. [14]

    Kong, J.; Kim, J.; and Bae, J. 2020. Hifi-gan: Generative adversarial networks for efficient and high fidelity speech synthesis. Advances in neural information processing systems, 33: 17022--17033

    2020

  7. [15]

    Lee, Y.; Shin, J.; and Jung, K. 2020. Bidirectional variational inference for non-autoregressive text-to-speech. In International conference on learning representations

    2020

  8. [16]

    Liu, J.; Li, C.; Ren, Y.; Chen, F.; and Zhao, Z. 2022. Diffsinger: Singing voice synthesis via shallow diffusion mechanism. In Proceedings of the AAAI conference on artificial intelligence, volume 36, 11020--11028

    2022

  9. [17]

    Lu, P.; Wu, J.; Luan, J.; Tan, X.; and Zhou, L. 2020. Xiaoicesing: A high-quality and integrated singing voice synthesis system. arXiv preprint arXiv:2006.06261

    2020 arXiv

  10. [18]

    Ma, X.; Zhou, C.; Li, X.; Neubig, G.; and Hovy, E. 2019. F low S eq: Non-Autoregressive Conditional Sequence Generation with Generative Flow. In Inui, K.; Jiang, J.; Ng, V.; and Wan, X., eds., Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processi...

    2019

  11. [19]

    B.; Oliverio, J.; and Clements, M

    Macon, M.; Jensen-Link, L.; George, E. B.; Oliverio, J.; and Clements, M. 1997. Concatenation-based midi-to-singing voice synthesis. In Audio Engineering Society Convention 103. Audio Engineering Society

    1997

  12. [20]

    McAuliffe, M.; Socolof, M.; Mihuc, S.; Wagner, M.; and Sonderegger, M. 2017. Montreal forced aligner: Trainable text-speech alignment using kaldi. In Interspeech, volume 2017, 498--502

    2017

  13. [21]

    Nakamura, K.; Hashimoto, K.; Oura, K.; Nankaku, Y.; and Tokuda, K. 2019. Singing voice synthesis based on convolutional neural networks. arXiv preprint arXiv:1904.06868

    2019 arXiv

  14. [22]

    Nishimura, M.; Hashimoto, K.; Oura, K.; Nankaku, Y.; and Tokuda, K. 2016. Singing Voice Synthesis Based on Deep Neural Networks. In Interspeech, 2478--2482

    2016

  15. [23]

    Oord, A. v. d.; Dieleman, S.; Zen, H.; Simonyan, K.; Vinyals, O.; Graves, A.; Kalchbrenner, N.; Senior, A.; and Kavukcuoglu, K. 2016. Wavenet: A generative model for raw audio. arXiv preprint arXiv:1609.03499

    2016 arXiv

  16. [24]

    Ren, Y.; Ruan, Y.; Tan, X.; Qin, T.; Zhao, S.; Zhao, Z.; and Liu, T.-Y. 2019. Fastspeech: Fast, robust and controllable text to speech. Advances in neural information processing systems, 32

    2019

  17. [25]

    Saino, K.; Zen, H.; Nankaku, Y.; Lee, A.; and Tokuda, K. 2006. An HMM-based singing voice synthesis system. In Ninth International Conference on Spoken Language Processing

    2006

  18. [26]

    Schuster, M.; and Paliwal, K. K. 1997. Bidirectional recurrent neural networks. IEEE transactions on Signal Processing, 45(11): 2673--2681

    1997

  19. [27]

    J.; Schuster, M.; Jaitly, N.; Yang, Z.; Chen, Z.; Zhang, Y.; Wang, Y.; Skerrv-Ryan, R.; et al

    Shen, J.; Pang, R.; Weiss, R. J.; Schuster, M.; Jaitly, N.; Yang, Z.; Chen, Z.; Zhang, Y.; Wang, Y.; Skerrv-Ryan, R.; et al. 2018. Natural tts synthesis by conditioning wavenet on mel spectrogram predictions. In 2018 IEEE international conference on acoustics, speech and signa...

    2018

  20. [28]

    Shi, Y.; Wang, Y.; Wu, C.; Yeh, C.-F.; Chan, J.; Zhang, F.; Le, D.; and Seltzer, M. 2021. Emformer: Efficient memory transformer based acoustic model for low latency streaming speech recognition. In ICASSP 2021-2021 IEEE International Conference on Acoustics, Speech and Signal...

    2021

  21. [29]

    N.; Kaiser, .; and Polosukhin, I

    Vaswani, A.; Shazeer, N.; Parmar, N.; Uszkoreit, J.; Jones, L.; Gomez, A. N.; Kaiser, .; and Polosukhin, I. 2017. Attention is all you need. Advances in neural information processing systems, 30

    2017

  22. [30]

    J.; Jaitly, N.; Yang, Z.; Xiao, Y.; Chen, Z.; Bengio, S.; et al

    Wang, Y.; Skerry-Ryan, R.; Stanton, D.; Wu, Y.; Weiss, R. J.; Jaitly, N.; Yang, Z.; Xiao, Y.; Chen, Z.; Bengio, S.; et al. 2017. Tacotron: Towards end-to-end speech synthesis. arXiv preprint arXiv:1703.10135

    2017 arXiv

  23. [31]

    Wang, Y.; Wang, X.; Zhu, P.; Wu, J.; Li, H.; Xue, H.; Zhang, Y.; Xie, L.; and Bi, M. 2022. Opencpop: A high-quality open source chinese popular song corpus for singing voice synthesis. arXiv preprint arXiv:2201.07429

    2022 arXiv

  24. [32]

    D.; Marković, D.; and Richard, A

    Wu, Y.-C.; Gebru, I. D.; Marković, D.; and Richard, A. 2023. A udio D ec: An Open-Source Streaming High-Fidelity Neural Audio Codec. In ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

    2023

  25. [33]

    Yoshimura, T.; Fujimoto, T.; Oura, K.; and Tokuda, K. 2023. SPTK4 : An open-source software toolkit for speech signal processing. In 12th ISCA Speech Synthesis Workshop (SSW 2023), 211--217

    2023

  26. [34]

    Zhang, Y.; Cong, J.; Xue, H.; Xie, L.; Zhu, P.; and Bi, M. 2022. Visinger: Variational inference with adversarial learning for end-to-end singing voice synthesis. In ICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 7237--7241. IEEE

    2022

  27. [35]

    Zhang, Y.; Xue, H.; Li, H.; Xie, L.; Guo, T.; Zhang, R.; and Gong, C. 2023. VISinger2: High-Fidelity End-to-End Singing Voice Synthesis Enhanced by Digital Signal Processing Synthesizer . In Proc. INTERSPEECH 2023, 4444--4448

    2023

  28. [36]

    Zhang, Y.-J.; Pan, S.; He, L.; and Ling, Z.-H. 2019. Learning latent representations for style control and transfer in end-to-end speech synthesis. In ICASSP 2019-2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 6945--6949. IEEE

    2019

  29. [37]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all...

  30. [38]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.