mlsanityData to 14 Oct 2026

Mitigating data replication in text-to-audio generative diffusion models through anti-memorization guidance

arXiv 2509.14934 ↗First posted 18 Sep 2025eess.AS cs.LG cs.SDcited by 2

On the map: Diffusion models → Audio, speech and music

Abstract (arXiv)

A persistent challenge in generative audio models is data replication, where the model unintentionally generates parts of its training data during inference. In this work, we address this issue in text-to-audio diffusion models by exploring the use of anti-memorization strategies. We adopt Anti-Memorization Guidance (AMG), a technique that modifies the sampling process of pre-trained diffusion models to discourage memorization. Our study explores three types of guidance within AMG, each designed to reduce replication while preserving generation quality. We use Stable Audio Open as our backbone, leveraging its fully open-source architecture and training dataset. Our comprehensive experimental analysis suggests that AMG significantly mitigates memorization in diffusion-based text-to-audio generation without compromising audio fidelity or semantic alignment.

Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Placed in Audio, speech and music, one of 14 directions in Diffusion models. Builds on 26 of its 32 references that are in the map. First posted on arXiv 18 Sep 2025.

Builds on

26 of 32 references.

Closest ancestors · 4

Same direction, newest first

  1. 2025Diffused Responsibility: Analyzing the Energy Consumption of Generative Text-to-Audio Diffusion Models12 May
  2. 2024Tango 2: Aligning Diffusion-based Text-to-Audio Generations through Direct Preference Optimization15 Apr
  3. 2023Generation or Replication: Auscultating Audio Latent Diffusion Models16 Oct
  4. 2023AudioLDM 2: Learning Holistic Audio Generation with Self-supervised Pretraining10 Aug

Other references · 22

  1. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  2. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  3. 2022Scalable Diffusion Models with Transformerscited by 6,054
  4. 2022Classifier-Free Diffusion Guidancecited by 5,847
  5. 2023Extracting Training Data from Diffusion Modelscited by 732
  6. 2022Large-scale Contrastive Language-Audio Pretraining with Feature Fusion and Keyword-to-Caption Augmentationcited by 719
  7. 2023Simple and Controllable Music Generationcited by 488
  8. 2023MusicLM: Generating Music From Textcited by 448
  9. 2022Diffusion Art or Digital Forgery? Investigating Data Replication in Diffusion Modelscited by 424
  10. 2023Ablating Concepts in Text-to-Image Diffusion Modelscited by 317
  11. 2018Fr'echet Audio Distance: A Metric for Evaluating Music Enhancement Algorithmscited by 225
  12. 2023Understanding and Mitigating Copying in Diffusion Modelscited by 209
10 more
  1. 2016FMA: A Dataset For Music Analysiscited by 207
  2. 2024Stable Audio Opencited by 197
  3. 2023MERT: Acoustic Music Understanding Model with Large-Scale Self-supervised Trainingcited by 181
  4. 2024Detecting, Explaining, and Mitigating Memorization in Diffusion Modelscited by 95
  5. 2023Adapting Frechet Audio Distance for Generative Music Evaluationcited by 92
  6. 2024Towards Memorization-Free Diffusion Modelscited by 50
  7. 2025KAD: No More FAD! An Effective and Efficient Evaluation Metric for Audio Generationcited by 24
  8. 2025Aligning Text-to-Music Evaluation with Human Preferencescited by 21
  9. 2024Towards Assessing Data Replication in Music Generation with Music Similarity Metrics on Raw Audiocited by 10
  10. 2024PAGURI: a user experience study of creative interaction with text-to-music modelscited by 4

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 1; 2026: 1.

2 papers.

  1. 2025AI-Assisted Music Production: A User Study on Text-to-Music Modelscited by 1
  2. 2026Ghost in the Encoder: Decodable Artist Identity Representations in Lyrics-to-Song Generationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Stage-adaptive audio diffusion modeling
  2. 2023ConsistencyTTA: Accelerating Diffusion-Based Text-to-Audio Generation with Consistency Distillation
  3. 2026SoundWeaver: Semantic Warm-Starting for Text-to-Audio Diffusion Serving
  4. 2026Characterizing Memorization in Diffusion Language Models: Generalized Extraction and Sampling Effects
  5. 2024PTQ4ADM: Post-Training Quantization for Efficient Text Conditional Audio Diffusion Models
  6. 2024Auffusion: Leveraging the Power of Diffusion and Large Language Models for Text-to-Audio Generation
  7. 2026UAT: Unified Audio-Text Diffusion for Audio Generation, Editing, and Captioning
  8. 2025ControlAudio: Tackling Text-Guided, Timing-Indicated and Intelligible Audio Generation via Progressive Diffusion Modeling