Pith. sign in

REVIEW 4 major objections 5 minor 29 references

SILA: Signal-to-Language Augmentation for Enhanced Control in Text-to-Audio Generation

T0 review · 4 major / 5 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 This paper claims that appending automatically extracted acoustic tags to text captions during training can give any text-conditioned audio generation model fine-grained control over loudness, pitch, reverb, fade, brightness, noise, and…

desk verdict SILA is a clean caption-augmentation recipe for TTA control; the subjective evidence is real, but the disentanglement claim outruns the metrics, and the objective check for three descriptors is circular. read the letter →

arxiv 2412.09789 v1 pith:GDKC2K7Y submitted 2024-12-13 cs.SD eess.AS

Sonal Kumar , Prem Seetharaman , Justin Salamon , Dinesh Manocha , Oriol Nieto This is my paper · ORCID
classification cs.SDeess.AS
keywords text-to-audiogenerationacousticdescriptorsfine-grainedcontrolsignal-to-languageaugmentationdiffusionmodelssoundeffectsdisentangledrepresentationscaption
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

SILA is a training-time recipe for text-to-audio models: before training, every sound in the dataset gets a caption, and attached to that caption are short labels describing measurable acoustic properties such as loudness band, pitch range, reverb level, fade direction, brightness, background noise, and duration. At inference the user can write or edit those same labels, and the model learns to treat them as separate control knobs rather than as part of the scene description. The paper's central claim is that this augmentation teaches the model disentangled representations of acoustic characteristics, so generated audio follows the requested property without the overall semantic alignment or perceptual quality collapsing. On their own text-to-audio diffusion transformer, SILA improves the CLAP text-audio alignment score over the no-augmentation baseline and over Stable Audio Open, AudioGen, and Tango 2, with a competitive FAD, and listeners preferred SILA on every evaluated attribute. The method is deliberately model-agnostic, so the same trick should carry over to other text-conditioned audio generators.

What carries the argument

The machinery is the SILA caption: a normal semantic caption followed by a block of '& descriptor: value' tags, such as '& loudness: soft, & pitch: low, & reverb: very wet'. The descriptor values are produced by small signal estimators—loudness in LKFS bands, pitch in octave ranges via a neural pitch tracker, brightness via spectral centroid, noise via an SNR-style frame comparison—together with reverb and fade classes created by data augmentation and a duration label. Training a text-conditioned diffusion transformer on these concatenated strings is what forces the language conditioning to carry explicit acoustic information as separable dimensions, so that at inference a user can edit a tag and the generated audio changes only that property.

What would settle it

Generate the same prompt with only the loudness descriptor changed between 'very soft' and 'very loud', measure the integrated loudness of the outputs with an independent loudness meter, and have listeners rank them; if there is no consistent, perceptible level gap across many prompts, the control claim fails.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that fine-grained acoustic control in text-to-audio generation can be achieved without new architectures, auxiliary networks, or inference-time guidance: the model just needs to see the signal-level truth during training. The authors append descriptor strings such as '& loudness: soft' and '& reverb: very wet' to captions, where the descriptors come from signal estimates, simulated reverb and fade via data augmentation, and a duration label. Because the same label appears across many different semantic events, the text encoder can separate what the sound is from how loud, how reverberant, or how bright it is. The result claimed is higher CLAP alignment than all baselines, comparable FAD, and subjective preference across loudness, pitch, reverb, noise, fade, duration, and overall alignment.

Load-bearing premise

The approach assumes that the automatic labels describing each training sound—how loud, how high-pitched, how bright, how noisy—truly match what listeners perceive, since the model can only learn to control what its labels actually measure.

Editorial extensions

If this is right

  • Retraining an existing text-to-audio model on SILA-style captions should transfer the same descriptor vocabulary to new datasets, because the descriptors are computed from the audio itself rather than requiring human annotations.
  • A user at inference can specify 'very loud', 'very wet', 'bright', or 'fade out' in the prompt and expect the generated sound to land in the corresponding measured range, as the paper's objective results indicate.
  • The soft versus loud distinction is learned as a contextual difference, not a volume knob: a soft explosion sounds distant while a loud one sounds close.
  • The method is confined to single-event sound effects; compositional scenes and text-to-speech are explicitly left as limitations.

Reading between the lines

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

  • Because the descriptor vocabulary is categorical and estimate-based, SILA should be compatible with editor-style controls such as sliders, presets, or partial prompt editing, turning text prompts into a parametric surface without extra model machinery; the paper does not build such an interface.
  • The same training-time augmentation could extend to attributes the paper only lists as future work, such as stereo width, panning, and apparent source motion, since those are also measurable signal properties that could be phrased as descriptor tags.
  • An independent test of the descriptors with a second estimator or human labels would separate real perceptual control from control over artifacts of the chosen estimators; the paper's objective evaluation of noise, brightness, and pitch reuses the estimators that created the labels.
  • If the disentanglement story holds, editing or deleting one descriptor tag at inference should leave the other acoustic properties and the semantic content intact, which would make SILA a natural fit for prompt-to-prompt audio editing; this is a direct corollary the paper does not demonstrate.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper proposes SILA, a training-time augmentation for text-to-audio models: acoustic descriptors (loudness, pitch, reverb, noise, brightness, fade, duration) are computed from training audio and appended to the text caption as structured tokens such as '& loudness: soft'. A DiT-based text-to-audio model is trained on these augmented captions, and at inference the user writes the same descriptor format to control acoustic attributes. The method is evaluated on the AuditionSFX dataset against Stable Audio Open, AudioGen, and Tango 2, using CLAP score, FAD, objective acoustic descriptor values (Table II), and a subjective listening test with 22 participants (Table III). The paper reports the highest CLAP score, comparable FAD, and strong subjective preference for SILA.

Significance. If the central claim holds, SILA is a simple, model-agnostic recipe for adding fine-grained acoustic control to text-to-audio models without architectural changes, which would be useful for sound design and creative production. The paper has several concrete strengths: the audio examples are available on a project page; the subjective test covers six descriptors plus overall alignment; reverb and duration are evaluated with metrics independent of the training-label estimators; and the CLAP/FAD results indicate that control does not obviously degrade generation quality. However, the load-bearing evidence for disentangled control is currently incomplete: three of the seven descriptors are evaluated with the same estimators used to build the training labels, no single-descriptor counterfactual experiment is reported, and no statistical uncertainty is provided for the subjective or objective comparisons. These issues are fixable and do not undermine the plausibility of the method, but they must be addressed before the central claims are established.

major comments (4)
  1. [Section V, Table II; Section III-B] The objective evaluation for noise, brightness, and pitch is circular. Table II reports these three rows using 'the metrics discussed in Section III-B,' i.e., the same SNR frame-contrast, spectral-centroid, and CREPE-octave estimators that generated the SILA training labels. The comparison therefore shows that the model reproduces its own labeler, not that these descriptors are perceptually meaningful or disentangled. Brightness is especially exposed because it has no column in the subjective study of Table III, so its entire support rests on this circular metric. Consequently, the only non-circular objective evidence is reverb (RT60) and duration, while loudness and fade have no objective evaluation at all.
  2. [Section V, soft explosion example] No counterfactual single-descriptor manipulation is reported. Disentangled control requires varying one descriptor token while holding the semantic caption and all other descriptor tokens fixed; Table II instead compares aggregate per-category means over different captions. The soft/loud explosion example is not such a test, because changing loudness also changes the plausible source size and distance ('soft explosion ... in the distance' vs 'loud explosion ... very close'), which could be a semantically sensible rendition rather than independent acoustic control. Please add within-caption, single-descriptor paired comparisons for each of the seven descriptors.
  3. [Tables II and III] No error bars, confidence intervals, or statistical tests are reported anywhere. The subjective results come from 22 participants and 30 items, and proportions such as SILA 0.36 vs AudioGen 0.22 for duration or SILA 0.50 vs Stable Audio 0.23 for pitch need paired significance tests before preference claims are warranted. Table II reports single average values (e.g., baseline 4.31 vs SILA 6.78 for silent-background SNR), and the CLAP/FAD gaps (0.29 vs 0.27; 0.84 vs 0.81) also lack variance estimates, so the reliability of all headline comparisons is unquantified.
  4. [Abstract and Section IV-C] The 'model-agnostic' claim is not empirically supported by the experiments. All training and evaluation use a single DiT-based text-to-audio model; the paper does not instantiate SILA with Stable Audio Open, AudioGen, Tango 2, or any other text-conditioned backbone. If the claimed contribution is that any text-conditioned model can be made controllable without architectural changes, at least one additional architecture should be tested.
minor comments (5)
  1. [Section III-B.1] The loudness classes leave an unlabeled gap between -40 and -30 LKFS; please state the rule for audio files falling in that gap, and whether they are excluded from the descriptor-labeled training subset.
  2. [Section III-B.7] The duration descriptor is said to be 'probabilistically appended' to the metadata, but the probability is not specified; please state the value and confirm whether the same probability is used at inference.
  3. [Section III-A] The model used for caption refinement is called both 'Mistral-7B' and 'Mixtral-7B'; please make the naming consistent.
  4. [Section IV, Training Datasets] The paper defers dataset statistics to the project page; including the number of training samples, the class distribution for each descriptor, and the augmentation proportions would make the training setup self-contained and reproducible.
  5. [Section IV, User-Evaluation] Please clarify whether each participant made one choice per category per trial or one overall choice, and how ties or incomplete responses were handled in the reported proportions.

Circularity Check

1 steps flagged · score 4.0 of 10

Objective noise/brightness/pitch evaluation reuses the Section III-B estimators used to create SILA labels; brightness rests entirely on that circular metric.

  1. self definitional [Section V, Results (Table II); Section III-B, Audio Descriptors]
    "For the comparison of noise, brightness, and pitch, we use the metrics discussed in Section III-B."

    Section III-B defines these descriptors with the same estimators used to build the training captions: pitch uses CREPE octave bands (low <1.5, high >3.5), noise uses Mel-spectrogram SNR thresholds (SNR≥6 silent, SNR≤2 noisy), and brightness uses spectral centroid thresholds (<45 dull, >65 bright). Table II then reports 'Noise (SNR)', 'Brightness (spec. centroid)', 'Pitch (octave)' using 'the metrics discussed in Section III-B' and concludes values are 'within the expected range'. Thus for these three descriptors, 'control' means matching the same classifier that wrote the training labels. This verifies reproduction of the labeling function, not independent perceptual or disentangled control. Brightness has no subjective column, so its support reduces entirely to this estimator.

full rationale

SILA's contribution is empirical rather than a formal derivation, so most of the paper is not circular in the strict sense: the model-agnostic augmentation pipeline can be implemented and tested independently, CLAP/FAD are external benchmarks, reverb is measured with an external RT60 timbral model, duration is measured in seconds, and the subjective study (Table III) independently covers loudness, pitch, reverb, noise, fade, duration, and overall alignment. The one load-bearing circular element is the objective evaluation of noise, brightness, and pitch in Table II, which reuses the Section III-B estimators that generated the training labels; for brightness this is the only support because Table III omits a brightness column. The self-citation of GAMA as the captioning backbone is not counted as circular: GAMA is a published external component used as a tool, not the target result, and it is not the source of the controllability claim. The 'soft explosion' example is a confound rather than circularity, and is not scored here.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The method introduces no new physical or conceptual entities. It relies on a stack of domain assumptions: LLM captioners produce faithful captions, DSP estimators capture perceptual categories, simulated effects generalize, and CLAP/FAD are valid metrics. The hand-chosen descriptor thresholds are the main free parameters, and they are not justified against an external standard.

free parameters (6)
  • Loudness classification boundaries = very soft -70 to -55 LKFS, soft -55 to -40 LKFS, loud -30 to -15 LKFS, very loud >-15 LKFS
    Hand-chosen in Section III-B.1 from the training data distribution; intermediate loudness files are excluded from categorization.
  • Pitch classification thresholds = low < 1.5 octaves, high > 3.5 octaves
    Hand-chosen in Section III-B.2 based on perceived frequency ranges; middle range is left uncategorized.
  • Brightness classification thresholds = spectral centroid less than 45 (dull), greater than 65 (bright)
    Set in Section III-B.5 based on the distribution of spectral centroid values in the training data; middle values are excluded.
  • SNR classification thresholds = silent background SNR >= 6, noisy background SNR <= 2
    Chosen in Section III-B.4 because the metric was 'somewhat unreliable' in the middle range; captions are not augmented for SNR between 2 and 6.
  • Reverb augmentation parameters = not reported
    Section III-B.3 says Pedalboard is used to create dry, slightly wet, wet, and very wet categories, but the specific reverb settings are not given.
  • Fade augmentation parameters = not reported
    Section III-B.6 says synthesized fade-in and fade-out transitions are added, but ramp lengths and shapes are not specified.
assumptions (5)
  • domain assumption GAMA and Mistral-7B generate semantically accurate coarse captions for each audio clip.
    Section III-A uses these captions as the base text before appending descriptors; hallucinated or mismatched captions would corrupt the training signal.
  • domain assumption CREPE pitch estimation, LKFS loudness, spectral centroid, and the SNR estimate are reliable proxies for the perceptual categories low/high, soft/loud, dull/bright, and silent/noisy.
    Section III-B defines categories from these estimators, and the objective evaluation in Section V reuses them as ground truth.
  • domain assumption Simulated reverb and fade effects applied with Pedalboard approximate real-world reverb and fade well enough for the model to generalize.
    Section III-B.3 and III-B.6 augment training audio with synthetic effects because natural examples are lacking.
  • domain assumption CLAP score and FAD are valid measures of text-audio alignment and generation plausibility for this task.
    Section IV uses these as the objective metrics without calibration or error analysis.
  • domain assumption The DiT model with FLAN-T5 conditioning can learn associations between the appended descriptor tokens and the corresponding acoustic output properties.
    Section IV assumes the architecture can exploit augmented captions; this is the central learning assumption underlying the claimed disentanglement.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SILA: Signal-to-Language Augmentation for Enhanced Control in Text-to-Audio Generation." pith.science (2026). https://pith.science/paper/GDKC2K7Y

@misc{pith2026241209789,
  author       = {Pith},
  title        = {Pith review of: SILA: Signal-to-Language Augmentation for Enhanced Control in Text-to-Audio Generation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GDKC2K7Y}},
  note         = {Machine review of arXiv:2412.09789}
}
read the original abstract

The field of text-to-audio generation has seen significant advancements, and yet the ability to finely control the acoustic characteristics of generated audio remains under-explored. In this paper, we introduce a novel yet simple approach to generate sound effects with control over key acoustic parameters such as loudness, pitch, reverb, fade, brightness, noise and duration, enabling creative applications in sound design and content creation. These parameters extend beyond traditional Digital Signal Processing (DSP) techniques, incorporating learned representations that capture the subtleties of how sound characteristics can be shaped in context, enabling a richer and more nuanced control over the generated audio. Our approach is model-agnostic and is based on learning the disentanglement between audio semantics and its acoustic features. Our approach not only enhances the versatility and expressiveness of text-to-audio generation but also opens new avenues for creative audio production and sound design. Our objective and subjective evaluation results demonstrate the effectiveness of our approach in producing high-quality, customizable audio outputs that align closely with user specifications.

Figures

Figures reproduced from arXiv: 2412.09789 by the authors.

Figure 1
Figure 1. Illustration of our proposed methodology for training (left) and inference using (right) SILA. During training, the augmentation for reverb and fade is not always applied to the audio, and is only used for augmenting the audios to add reverb and fade descriptors. III. METHOD [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

29 extracted references · 15 canonical work pages

  1. [1]

    Stable audio open,

    Zach Evans, Julian D Parker, CJ Carr, Zack Zukowski, Josiah Taylor, and Jordi Pons, “Stable audio open,” arXiv preprint arXiv:2407.14358 , 2024

    arXiv 2024

  2. [2]

    AudioLDM: Text-to- audio generation with latent diffusion models,

    Haohe Liu, Zehua Chen, Yi Yuan, Xinhao Mei, Xubo Liu, Danilo Mandic, Wenwu Wang, and Mark D Plumbley, “AudioLDM: Text-to- audio generation with latent diffusion models,” in Proceedings of the 40th International Conference on Machine Learning , Andreas Krause, Emma Brunskill, Kyunghyun Cho, Barbara Engelhardt, Sivan Sabato, and Jonathan Scarlett, Eds. 23–2...

    work page 2023

  3. [3]

    Denoising diffusion prob- abilistic models,

    Jonathan Ho, Ajay Jain, and Pieter Abbeel, “Denoising diffusion prob- abilistic models,” Advances in neural information processing systems , vol. 33, pp. 6840–6851, 2020

    2020

  4. [4]

    Neural codec language models are zero-shot text to speech synthesiz- ers,

    Chengyi Wang, Sanyuan Chen, Yu Wu, Ziqiang Zhang, Long Zhou, Shujie Liu, Zhuo Chen, Yanqing Liu, Huaming Wang, Jinyu Li, et al., “Neural codec language models are zero-shot text to speech synthesiz- ers,” arXiv preprint arXiv:2301.02111 , 2023

    arXiv 2023

  5. [5]

    Simple and controllable music generation,

    Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi, and Alexandre D´efossez, “Simple and controllable music generation,” Advances in Neural Information Processing Systems, vol. 36, 2024

    2024

  6. [6]

    Compa: Addressing the gap in compositional reasoning in audio-language models,

    Sreyan Ghosh, Ashish Seth, Sonal Kumar, Utkarsh Tyagi, Chandra Kiran Reddy Evuru, Ramaneswaran S, S Sakshi, Oriol Nieto, Ramani Duraiswami, and Dinesh Manocha, “Compa: Addressing the gap in compositional reasoning in audio-language models,” in The Twelfth International Conference on Learning Representations , 2024

    work page 2024

  7. [7]

    A Demand-Driven Perspective on Generative Audio AI

    Sangshin Oh, Minsung Kang, Hyeongi Moon, Keunwoo Choi, and Ben Sangbae Chon, “A demand-driven perspective on generative audio ai,” arXiv preprint arXiv:2307.04292 , 2023

    work page Pith review arXiv 2023

  8. [8]

    Large-scale contrastive language- audio pretraining with feature fusion and keyword-to-caption augmen- tation,

    Yusong Wu, Ke Chen, Tianyu Zhang, Yuchen Hui, Taylor Berg- Kirkpatrick, and Shlomo Dubnov, “Large-scale contrastive language- audio pretraining with feature fusion and keyword-to-caption augmen- tation,” in ICASSP 2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . IEEE, 2023, pp. 1–5

    work page 2023

Show all 29 references
  1. [9]

    Fr \’echet audio distance: A metric for evaluating music enhance- ment algorithms,

    Kevin Kilgour, Mauricio Zuluaga, Dominik Roblek, and Matthew Shar- ifi, “Fr \’echet audio distance: A metric for evaluating music enhance- ment algorithms,” arXiv preprint arXiv:1812.08466 , 2018

    2018 arXiv

  2. [10]

    Generative adversarial nets,

    Ian J. Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron C. Courville, and Yoshua Bengio, “Generative adversarial nets,” in Neural Information Processing Systems, 2014

    2014

  3. [11]

    Tacotron: Towards end-to-end speech synthesis,

    Yuxuan Wang, RJ Skerry-Ryan, Daisy Stanton, Yonghui Wu, Ron J Weiss, Navdeep Jaitly, Zongheng Yang, Ying Xiao, Zhifeng Chen, Samy Bengio, et al., “Tacotron: Towards end-to-end speech synthesis,” arXiv preprint arXiv:1703.10135, 2017

    2017 arXiv

  4. [12]

    Auto-encoding variational bayes,

    Diederik P Kingma, “Auto-encoding variational bayes,” arXiv preprint arXiv:1312.6114, 2013

    2013 arXiv

  5. [13]

    Audiogen: Textually guided audio generation,

    Felix Kreuk, Gabriel Synnaeve, Adam Polyak, Uriel Singer, Alexan- dre D ´efossez, Jade Copet, Devi Parikh, Yaniv Taigman, and Yossi Adi, “Audiogen: Textually guided audio generation,” arXiv preprint arXiv:2209.15352, 2022

    2022 arXiv

  6. [14]

    Au- toregressive diffusion transformer for text-to-speech synthesis,

    Zhijun Liu, Shuai Wang, Sho Inoue, Qibing Bai, and Haizhou Li, “Au- toregressive diffusion transformer for text-to-speech synthesis,” arXiv preprint arXiv:2406.05551, 2024

    2024 arXiv

  7. [15]

    Diffwave: A versatile diffusion model for audio synthesis,

    Zhifeng Kong, Wei Ping, Jiaji Huang, Kexin Zhao, and Bryan Catan- zaro, “Diffwave: A versatile diffusion model for audio synthesis,” in International Conference on Learning Representations , 2021

    2021

  8. [16]

    Music controlnet: Multiple time-varying controls for music generation,

    Shih-Lun Wu, Chris Donahue, Shinji Watanabe, and Nicholas J Bryan, “Music controlnet: Multiple time-varying controls for music generation,” IEEE/ACM Transactions on Audio, Speech, and Language Processing , vol. 32, pp. 2692–2703, 2024

    2024

  9. [17]

    Hierarchical generative modeling for controllable speech synthesis,

    Wei-Ning Hsu, Yu Zhang, Ron J Weiss, Heiga Zen, Yonghui Wu, Yuxuan Wang, Yuan Cao, Ye Jia, Zhifeng Chen, Jonathan Shen, et al., “Hierarchical generative modeling for controllable speech synthesis,” arXiv preprint arXiv:1810.07217 , 2018

    arXiv

  10. [18]

    Deep voice: Real-time neural text-to-speech,

    Sercan ¨O. Arık, Mike Chrzanowski, Adam Coates, Gregory Diamos, Andrew Gibiansky, Yongguo Kang, Xian Li, John Miller, Andrew Ng, Jonathan Raiman, Shubho Sengupta, and Mohammad Shoeybi, “Deep voice: Real-time neural text-to-speech,” in Proceedings of the 34th International Conf...

    2017

  11. [19]

    GAMA: A large audio-language model with advanced audio understanding and complex reasoning abilities,

    Sreyan Ghosh, Sonal Kumar, Ashish Seth, Chandra Kiran Reddy Evuru, Utkarsh Tyagi, S Sakshi, Oriol Nieto, Ramani Duraiswami, and Dinesh Manocha, “GAMA: A large audio-language model with advanced audio understanding and complex reasoning abilities,” in Proceedings of the 2024 Co...

    2024

  12. [20]

    Mistral 7b,

    Albert Q Jiang, Alexandre Sablayrolles, Arthur Mensch, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Florian Bressand, Gianna Lengyel, Guillaume Lample, Lucile Saulnier, et al., “Mistral 7b,” arXiv preprint arXiv:2310.06825, 2023

    arXiv

  13. [21]

    Brian CJ Moore, An introduction to the psychology of hearing , Brill, 2012

    2012

  14. [22]

    Crepe: A convolutional representation for pitch estimation,

    Jong Wook Kim, Justin Salamon, Peter Li, and Juan Pablo Bello, “Crepe: A convolutional representation for pitch estimation,” in 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2018, pp. 161–165

    2018

  15. [23]

    Suppression of acoustic noise in speech using spectral subtraction,

    S. Boll, “Suppression of acoustic noise in speech using spectral subtraction,” IEEE Transactions on Acoustics, Speech, and Signal Processing, vol. 27, no. 2, pp. 113–120, 1979

    1979

  16. [24]

    Fsd50k: an open dataset of human-labeled sound events,

    Eduardo Fonseca, Xavier Favory, Jordi Pons, Frederic Font, and Xavier Serra, “Fsd50k: an open dataset of human-labeled sound events,” IEEE/ACM Transactions on Audio, Speech, and Language Processing , vol. 30, pp. 829–852, 2021

    2021

  17. [25]

    Make- an-audio: Text-to-audio generation with prompt-enhanced diffusion mod- els,

    Rongjie Huang, Jiawei Huang, Dongchao Yang, Yi Ren, Luping Liu, Mingze Li, Zhenhui Ye, Jinglin Liu, Xiang Yin, and Zhou Zhao, “Make- an-audio: Text-to-audio generation with prompt-enhanced diffusion mod- els,” in International Conference on Machine Learning . PMLR, 2023, pp. 1...

    2023

  18. [26]

    Scalable diffusion models with transformers,

    William Peebles and Saining Xie, “Scalable diffusion models with transformers,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 4195–4205

    2023

  19. [27]

    Scaling instruction-finetuned language models,

    Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Yunxuan Li, Xuezhi Wang, Mostafa Dehghani, Sid- dhartha Brahma, et al., “Scaling instruction-finetuned language models,” Journal of Machine Learning Research, vol. 25, no. 70, pp. 1–53, 2024

    2024

  20. [28]

    High-fidelity audio compression with improved rvqgan,

    Rithesh Kumar, Prem Seetharaman, Alejandro Luebs, Ishaan Kumar, and Kundan Kumar, “High-fidelity audio compression with improved rvqgan,” in Proceedings of the 37th International Conference on Neural Information Processing Systems, Red Hook, NY , USA, 2024, NIPS ’23, Curran As...

    2024

  21. [29]

    Tango 2: Aligning diffusion-based text-to-audio generations through direct preference optimization,

    Navonil Majumder, Chia-Yu Hung, Deepanway Ghosal, Wei-Ning Hsu, Rada Mihalcea, and Soujanya Poria, “Tango 2: Aligning diffusion-based text-to-audio generations through direct preference optimization,” in Proceedings of the 32nd ACM International Conference on Multimedia , 2024...

    2024

Pith tools

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