Scaling Transformers for Low-Bitrate High-Quality Speech Coding

• URL: http://arxiv.org/abs/2411.19842v1
• Date: Fri, 29 Nov 2024 16:58:02 GMT
• Title: Scaling Transformers for Low-Bitrate High-Quality Speech Coding
• Authors: Julian D Parker, Anton Smirnov, Jordi Pons, CJ Carr, Zack Zukowski, Zach Evans, Xubo Liu,
• Abstract summary: We show that it is possible to reach state-of-the-art speech quality at extremely low bit-rates of $400$ or $700$ bits-per-second.<n>The trained models strongly out-perform existing baselines in both objective and subjective tests.
• Score: 15.58137711465863
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: The tokenization of speech with neural audio codec models is a vital part of modern AI pipelines for the generation or understanding of speech, alone or in a multimodal context. Traditionally such tokenization models have concentrated on low parameter-count architectures using only components with strong inductive biases. In this work we show that by scaling a transformer architecture with large parameter count to this problem, and applying a flexible Finite Scalar Quantization (FSQ) based bottleneck, it is possible to reach state-of-the-art speech quality at extremely low bit-rates of $400$ or $700$ bits-per-second. The trained models strongly out-perform existing baselines in both objective and subjective tests.

Related papers

• One Quantizer is Enough: Toward a Lightweight Audio Codec [10.903708510237875]
  We present SQCodec, a lightweight neural audio that leverages a single quantizer to address limitations of existing approaches. SQCodec explores streamlined convolutional networks and local Transformer modules, alongside TConv. Experiments show that SQCodec audio quality comparable to multi-quantizer achieves baselines, while its single-quantizer design offers enhanced adaptability.
  arXiv  Detail & Related papers  (2025-04-07T11:34:39Z)
• Efficient Language Modeling for Low-Resource Settings with Hybrid RNN-Transformer Architectures [8.442206285783463]
  Transformer-based language models have recently been at the forefront of active research in text generation. These models' advances come at the price of prohibitive training costs, with parameter counts in the billions and compute requirements measured in petaflop/s-decades. We investigate transformer-based architectures for improving model performance in a low-data regime by selectively replacing attention layers with feed-forward and quasi-recurrent neural network layers.
  arXiv  Detail & Related papers  (2025-02-02T01:05:09Z)
• VQalAttent: a Transparent Speech Generation Pipeline based on Transformer-learned VQ-VAE Latent Space [0.49109372384514843]
  VQalAttent is a lightweight model designed to generate fake speech with tunable performance and interpretability. Our results demonstrate VQalAttent's capacity to generate intelligible speech samples with limited computational resources.
  arXiv  Detail & Related papers  (2024-11-22T00:21:39Z)
• Autoregressive Diffusion Transformer for Text-to-Speech Synthesis [39.32761051774537]
  We propose encoding audio as vector sequences in continuous space $mathbb Rd$ and autoregressively generating these sequences. High-bitrate continuous speech representation enables almost flawless reconstruction, allowing our model to achieve nearly perfect speech editing.
  arXiv  Detail & Related papers  (2024-06-08T18:57:13Z)
• High Fidelity Neural Audio Compression [92.4812002532009]
  We introduce a state-of-the-art real-time, high-fidelity, audio leveraging neural networks. It consists in a streaming encoder-decoder architecture with quantized latent space trained in an end-to-end fashion. We simplify and speed-up the training by using a single multiscale spectrogram adversary.
  arXiv  Detail & Related papers  (2022-10-24T17:52:02Z)
• ClusTR: Exploring Efficient Self-attention via Clustering for Vision Transformers [70.76313507550684]
  We propose a content-based sparse attention method, as an alternative to dense self-attention. Specifically, we cluster and then aggregate key and value tokens, as a content-based method of reducing the total token count. The resulting clustered-token sequence retains the semantic diversity of the original signal, but can be processed at a lower computational cost.
  arXiv  Detail & Related papers  (2022-08-28T04:18:27Z)
• FastLTS: Non-Autoregressive End-to-End Unconstrained Lip-to-Speech Synthesis [77.06890315052563]
  We propose FastLTS, a non-autoregressive end-to-end model which can directly synthesize high-quality speech audios from unconstrained talking videos with low latency. Experiments show that our model achieves $19.76times$ speedup for audio generation compared with the current autoregressive model on input sequences of 3 seconds.
  arXiv  Detail & Related papers  (2022-07-08T10:10:39Z)
• WavThruVec: Latent speech representation as intermediate features for neural speech synthesis [1.1470070927586016]
  WavThruVec is a two-stage architecture that resolves the bottleneck by using high-dimensional Wav2Vec 2.0 embeddings as intermediate speech representation. We show that the proposed model not only matches the quality of state-of-the-art neural models, but also presents useful properties enabling tasks like voice conversion or zero-shot synthesis.
  arXiv  Detail & Related papers  (2022-03-31T10:21:08Z)
• Sentence Bottleneck Autoencoders from Transformer Language Models [53.350633961266375]
  We build a sentence-level autoencoder from a pretrained, frozen transformer language model. We adapt the masked language modeling objective as a generative, denoising one, while only training a sentence bottleneck and a single-layer modified transformer decoder. We demonstrate that the sentence representations discovered by our model achieve better quality than previous methods that extract representations from pretrained transformers on text similarity tasks, style transfer, and single-sentence classification tasks in the GLUE benchmark, while using fewer parameters than large pretrained models.
  arXiv  Detail & Related papers  (2021-08-31T19:39:55Z)
• Revisiting Simple Neural Probabilistic Language Models [27.957834093475686]
  This paper revisits the neural probabilistic language model (NPLM) ofcitetBengio2003ANP. When scaled up to modern hardware, this model performs much better than expected on word-level language model benchmarks. Inspired by this result, we modify the Transformer by replacing its first self-attention layer with the NPLM's local concatenation layer.
  arXiv  Detail & Related papers  (2021-04-08T02:18:47Z)
• Any-to-Many Voice Conversion with Location-Relative Sequence-to-Sequence Modeling [61.351967629600594]
  This paper proposes an any-to-many location-relative, sequence-to-sequence (seq2seq), non-parallel voice conversion approach. In this approach, we combine a bottle-neck feature extractor (BNE) with a seq2seq synthesis module. Objective and subjective evaluations show that the proposed any-to-many approach has superior voice conversion performance in terms of both naturalness and speaker similarity.
  arXiv  Detail & Related papers  (2020-09-06T13:01:06Z)
• Pretraining Techniques for Sequence-to-Sequence Voice Conversion [57.65753150356411]
  Sequence-to-sequence (seq2seq) voice conversion (VC) models are attractive owing to their ability to convert prosody. We propose to transfer knowledge from other speech processing tasks where large-scale corpora are easily available, typically text-to-speech (TTS) and automatic speech recognition (ASR) We argue that VC models with such pretrained ASR or TTS model parameters can generate effective hidden representations for high-fidelity, highly intelligible converted speech.
  arXiv  Detail & Related papers  (2020-08-07T11:02:07Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.