mlsanityData to 14 Oct 2026

LILAC: An Idempotent Neural Speech Codec

arXiv 2608.05727 ↗First posted 6 Aug 2026cs.SD cs.LG eess.AScited by 0

Not on the map. Why?

Abstract (arXiv)

Neural Audio Codecs are widely adopted in speech generation and editing. However, existing neural audio codecs are not idempotent: across the paper's twelve baseline systems, every configuration tested rewrites, on average, at least 15% of its tokens in a single decode-re-encode pass. This poses a problem for utilizing Neural Audio Codecs as token interfaces in pipelines where re-encoding decoded outputs can occur. We present LILAC, a fully convolutional 24 kHz speech codec at 9.375 Hz and 0.75 kbit/s that is codec idempotent by construction; re-encoding the decoded audio of any valid token stream returns the identical stream. LILAC achieves idempotency while maintaining competitive quality, reaching UTMOS 4.14 and 4.24 on LibriSpeech and LibriTTS-R test sets, comparable to SOTA sub-1 kbit/s Neural Audio Codecs.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 21 of its 29 references that are in the map. First posted on arXiv 6 Aug 2026.

Builds on

21 of 29 references.

References · 21

  1. 2022Robust Speech Recognition via Large-Scale Weak Supervisioncited by 4,057
  2. 2020Reformer: The Efficient Transformercited by 1,668
  3. 2021WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processingcited by 1,657
  4. 2018Glow: Generative Flow with Invertible 1x1 Convolutionscited by 1,545
  5. 2019Normalizing Flows for Probabilistic Modeling and Inferencecited by 1,323
  6. 2018Averaging Weights Leads to Wider Optima and Better Generalizationcited by 1,209
  7. 2020HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesiscited by 1,154
  8. 2014NICE: Non-linear Independent Components Estimationcited by 1,058
  9. 2021SoundStream: An End-to-End Neural Audio Codeccited by 771
  10. 2023Finite Scalar Quantization: VQ-VAE Made Simplecited by 440
  11. 2023High-Fidelity Audio Compression with Improved RVQGANcited by 431
  12. 2018SDR - half-baked or well done?cited by 423
9 more
  1. 2022UTMOS: UTokyo-SaruLab System for VoiceMOS Challenge 2022cited by 390
  2. 2017The Reversible Residual Network: Backpropagation Without Storing Activationscited by 217
  3. 2024WavTokenizer: an Efficient Acoustic Discrete Codec Tokenizer for Audio Language Modelingcited by 129
  4. 2024Image and Video Tokenization with Binary Spherical Quantizationcited by 112
  5. 2023LibriTTS-R: A Restored Multi-Speaker Text-to-Speech Corpuscited by 101
  6. 2024SCOREQ: Speech Quality Assessment with Contrastive Regressioncited by 24
  7. 2025FocalCodec: Low-Bitrate Speech Coding via Focal Modulation Networkscited by 18
  8. 2025HiFiTTS-2: A Large-Scale High Bandwidth Speech Datasetcited by 10
  9. 2024Code Drift: Towards Idempotent Neural Audio Codecscited by 5

Built on by

Not cited yet.