Research map / Image generation / Fast diffusion sampling
PriorGrad: Improving Conditional Denoising Diffusion Models with Data-Dependent Adaptive Prior
Sang-gil Lee, Heeseung Kim, Chaehun Shin, Xu Tan, Chang Liu, Qi Meng, Tao Qin, Wei Chen, Sungroh Yoon, Tie-Yan Liu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use a conditional data-dependent Gaussian prior | DiffWave vocoder · LJSpeech test | LS-MAE: 0.5264 -> 0.5048 | Improved |
| Use a conditional data-dependent Gaussian prior | DiffWave vocoder · LJSpeech test | MR-STFT: 1.0920 -> 0.9976 | Improved |
| Use a conditional data-dependent Gaussian prior | DiffWave vocoder · LJSpeech test | MOS: 4.12 ± 0.08 -> 4.25 ± 0.08 | Improved |
| Use a conditional data-dependent Gaussian prior | DiffWave vocoder · LJSpeech test, 1M steps, 1.23M parameters | MOS: 3.90 ± 0.09 -> 4.02 ± 0.08 | Improved |
| Use phoneme-level data-dependent Gaussian prior | Diffusion acoustic model · LJSpeech acoustic model, 60K steps, 10M decoder parameters | MOS: 3.84 ± 0.10 -> 4.04 ± 0.07 | Improved |
| Use phoneme-level data-dependent Gaussian prior | Diffusion acoustic model · LJSpeech acoustic model, T_infer=2 | MOS: 2.80 ± 0.17 -> 4.25 ± 0.08 | Improved |
| Reduce decoder capacity from 10M to 3.5M parameters while retaining the PriorGrad prior (ablation) | PriorGrad acoustic model · LJSpeech acoustic model, 300K steps | MOS: 4.09 ± 0.08 -> 4.08 ± 0.07 | No clear effect |
| Replace the fixed data-dependent prior with a jointly trainable Gaussian prior (ablation) | PriorGrad acoustic model · LJSpeech acoustic model, 300K steps, 10M decoder parameters | MOS: 4.09 ± 0.08 -> 3.32 ± 0.12 | Hurt |
| Use voiced/unvoiced labels instead of frame-level spectral energy to construct the vocoder prior (ablation) | PriorGrad vocoder · LJSpeech | MOS | Hurt |
| Use phoneme-level statistics instead of frame-level spectral energy to construct the vocoder prior (ablation) | PriorGrad vocoder · LJSpeech | MOS | Hurt |
| Use T_infer=6 instead of T_infer=50 (ablation) | PriorGrad vocoder · LJSpeech test | MOS: 4.33 ± 0.07 -> 4.20 ± 0.08 | Hurt |
| Use T_infer=6 instead of T_infer=50 (ablation) | PriorGrad acoustic model · LJSpeech acoustic model | MOS: 4.39 ± 0.08 -> 4.29 ± 0.07 | Hurt |
Later papers that compared with this method
4 later accepted papers report PriorGrad as a baseline in their results tables; 4 of them report a better number on most of the results they share with it.
- RestoreGrad: Signal Restoration Using Conditional Denoising Diffusion Models with Jointly Learned Prior (ICML 2025 · better on 5 of 5 shared results)
- Cauchy Diffusion: A Heavy-tailed Denoising Diffusion Probabilistic Model for Speech Synthesis (AAAI 2025 · better on 2 of 2 shared results)
- RFWave: Multi-band Rectified Flow for Audio Waveform Reconstruction (ICLR 2025 · better on 1 of 1 shared result)
- From Discrete Tokens to High-Fidelity Audio Using Multi-Band Diffusion (NeurIPS 2023 · better on 1 of 1 shared result)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- DiffWave — DiffWave: A Versatile Diffusion Model for Audio Synthesis (ICLR 2021) · better on 8 of 8 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- RestoreGrad: Signal Restoration Using Conditional Denoising Diffusion Models with Jointly Learned Prior (ICML 2025 · similarity 0.90)
- Communication-Efficient Diffusion Denoising Parallelization via Reuse-then-Predict Mechanism (NeurIPS 2025 · similarity 0.89)
- Diffusion Models With Learned Adaptive Noise (NeurIPS 2024 · similarity 0.88)
- Diffusion Priors for Variational Likelihood Estimation and Image Denoising (NeurIPS 2024 · similarity 0.88)
- A Mixture-Based Framework for Guiding Diffusion Models (ICML 2025 · similarity 0.88)
- Information Theoretic Learning for Diffusion Models with Warm Start (NeurIPS 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
