---
title: Diffusion-Based Generative Models (EDMs)
url: https://www.emergentmind.com/topics/diffusion-based-generative-models-edms
type: topic
---

# Diffusion-Based Generative Models (EDMs)

Diffusion-based generative models (EDMs) define a powerful class of deep generative models rooted in forward–reverse stochastic processes, where a complex data distribution is mapped to noise via a prescribed noising process (typically a Markov chain or SDE), and generation is achieved by numerically approximating the reverse of this process using neural-parameterized drift or score fields. EDMs form the backbone of state-of-the-art algorithms for image, audio, graph, and scientific-data synthesis, and have been theoretically unified via connections to variational inference, score-matching, and energy-based modeling. Recent developments have advanced both the algorithmic efficiency and the theoretical foundation of these models, including non-asymptotic error bounds, accelerated sampling, energy-function parameterizations, and rigorous convergence analysis.

## 1. Probabilistic and Mathematical Foundations

EDMs are characterized by a two-stage architecture: a (non-learned) forward "diffusion" process and a (learned) reverse "denoising" process. In discrete time, with observed data $x_0\sim p_{\text{data}}(x_0)$ and a fixed schedule $\{\beta_t\}_{t=1}^T$, the forward process defines a Markov chain:

\[
q(x_t \mid x_{t-1}) = \mathcal{N}\left(x_t; \sqrt{1-\beta_t}\,x_{t-1},\,\beta_t I\right),
\quad \alpha_t=1-\beta_t,\,\bar{\alpha}_t=\prod_{i=1}^{t}\alpha_i,
\]
so that $q(x_t|x_0) = \mathcal{N}\left(\sqrt{\bar \alpha_t}\, x_0, (1-\bar \alpha_t)I\right)$ [2510.22230, 2306.09251, 2412.10948].

The continuous-time limit leads to an Itô SDE, e.g., the variance-preserving (VP) formulation:
\[
dx = -\tfrac{1}{2}\,\beta(t)\,x\,dt + \sqrt{\beta(t)}\,dW_t,
\]
with similar closed-form forward marginals [2404.09016, 2510.22230].

Reverse-time generative modeling uses another SDE, obtained via Anderson's theorem, with drift involving the unknown score $\nabla_x\log p_t(x)$ [2106.02808, 2404.09016]:
\[
dx = [f(x,t) - g(t)^2\,\nabla_x\log p_t(x)]dt + g(t)\,d\bar W_t,
\]
where the neural network $s_\theta(x,t)$ is trained to approximate the score [2412.10948]. The discrete reverse kernel, parameterized by a neural network as mean $\mu_\theta(x_t,t)$ and covariance $\Sigma_\theta(x_t,t)$, takes the general form:
\[
p_\theta(x_{t-1}|x_t) = \mathcal{N}\left(x_{t-1}; \mu_\theta(x_t,t), \Sigma_\theta(x_t,t)\right)
\]
or, in the widely used $\epsilon$-parameterization, neural networks directly predict the noise in the reverse process [2510.22230].

Key training objectives include the variational lower bound (ELBO), exact and denoising score matching (DSM), and their continuous-time analogues [2106.02808, 2404.09016, 2402.01965].

## 2. Training Objectives and Algorithmic Variants

The foundational objectives are:
- **Variational Lower Bound (ELBO)** on the marginal likelihood of the data:
  \[
  \mathcal{L}_{\text{VLB}} = \mathbb{E}[\text{KL}(q(x_T|x_0)\| p(x_T)) + \cdots + \text{KL}(q(x_{t-1}|x_t,x_0)\|p_\theta(x_{t-1}|x_t)) + \cdots]
  \]
  Each KL term may be written in closed-form for fixed (Gaussian) kernels [2412.10948].

- **Denoising Score Matching (DSM)**, fitting neural networks to conditional gradients of the noised data densities [2402.01965]:
  \[
  \ell(\theta;\sigma) = \mathbb{E}_{x_0,\tilde x\sim\mathcal{N}(x_0,\sigma^2I)}\left[\|\,s_\theta(\tilde x,\sigma) + \tfrac{\tilde x - x_0}{\sigma^2}\|_2^2\right].
  \]

- **Weighted Fisher Score Matching / Continuous-time ELBO** (for SDEs), connecting DSM to the negative log-density of the reverse-time process [2106.02808].

Algorithmic variants arise from changes to the objective or the generator–inference chain [2404.09016, 2510.22230]:
- **DDPM** (Ho et al.): vanilla ELBO-based discrete Gaussian diffusion.
- **Improved DDPM**: hybrid learned variance, cosine or other noise schedules.
- **Score-based models/NCSN**: DSM over a grid of noise levels, trained with annealed Langevin dynamics.
- **EDM/Elucidated Diffusion Model** (Karras et al.): continuous-time, noise-scale (σ) parameterization with high-order ODE samplers [2311.08667].
- **Energy-Based Diffusion Models (EBDMs)**: parameterize the reverse process via a neural scalar energy $E_\theta(x,t)$, enabling direct estimation of (unnormalized) log-priors and MH correction [2510.22230].

## 3. Sampling and Inference Mechanisms

Sampling proceeds by numerically integrating the learned reverse dynamics. The main algorithms include:
- **Ancestral Sampling**: backward Markov chain starting from $x_T \sim \mathcal{N}(0,I)$, with $x_{t-1}$ sampled from $p_\theta(x_{t-1}|x_t)$.
- **Probability-Flow ODE Solvers**: integrate deterministic ODEs corresponding to the SDE's marginal distributions (e.g., DDIM, DPM-Solver, exponential integrators), allowing for large step sizes and efficient sampling [2311.08667, 2210.12867, 2404.09016].
- **Langevin (Predictor–Corrector) Samplers**: alternate SDE-based predictor updates and score-based corrector (Langevin) steps [2404.09016, 2302.02591].
- **Metropolis–Hastings Corrected Diffusion**: in energy-based frameworks, each reverse transition is subjected to an MH test for bias/fidelity improvement [2510.22230].
- **Deep Equilibrium Solvers**: rephrase the entire DDIM chain as a single fixed-point system, allowing for parallel root-finding via Anderson acceleration, improving single-sample speed and inversion [2210.12867].
- **Discrete/Non-Gaussian and Constrained-domain Diffusion**: apply to categorical or structured data, e.g., via bridges or h-transforms [2208.14699, 2302.02591].

Representative sampling pseudocode for the stochastic DDPM sampler [2306.09251]:

```python
# DDPM sampling (stochastic)
Sample Y_T ∼ N(0,I)
for t = T…1:
    Y_{t−1} ← (Y_t + (1−α_t)s_t(Y_t))/√α_t + σ_t * Z_t
    Z_t ∼ N(0,I)
return Y₀
```

## 4. Theoretical Properties and Convergence

Recent research has provided non-asymptotic, finite-sample convergence rates for discrete-time diffusion samplers [2306.09251]. For a deterministic ODE-based sampler with $T$ steps and access to accurate scores:
- **Probability-flow ODE sampler**: Convergence in total-variation (TV) distance is $O(1/T)$, improving to $O(1/T^2)$ under further acceleration (bias/variance corrections).
- **Stochastic DDPM sampler**: TV (and KL) convergence rate is $O(1/\sqrt{T})$, boosted to $O(1/T)$ with variance correctors.

The TV bounds scale polynomially in data dimension $d$ and depend linearly on the mean squared error between the learned and the true score. No global smoothness or log-Sobolev assumptions are required; only boundedness of the forward process is necessary [2306.09251].

In energy-based diffusion models, explicit modeling of the log-prior energy enables true MH corrections, ensuring unbiased posterior sampling even under weak likelihood or strong prior regimes [2510.22230].

Variants based on bridge processes, e.g., for discrete or constrained domains, yield non-asymptotic KL divergence error bounds combining discretization and statistical estimation errors [2208.14699].

## 5. Practical Implementations and Applications

EDMs demonstrate broad versatility and have been instantiated in diverse domains:
- **Image, video, and audio synthesis**: EDMs trained in the raw or spectrogram domain (e.g., EDMSound) achieve state-of-the-art fidelity with accelerated ODE sampling (e.g., DPM-Solver), as demonstrated by Fréchet Audio Distance (FAD) and FID benchmarking [2311.08667].
- **Scientific and engineering inverse problems**: Energy-based EDMs with MH correction provide robust posterior sampling for high-dimensional parameter estimation in MIMO channel estimation, outperforming conventional DMs and other baselines in normalized MSE, even under limited pilot overhead [2510.22230].
- **Structured data, graphs, and molecules**: Graph-structured EDMs employ noise-perturbed adjacency matrices or node labels and learn permutation- and equivariant GNNs for score prediction, achieving SOTA in molecular conformation and design [2302.02591].
- **Materials science**: Denoising diffusion models reconstruct complex microstructures with minimal hand-engineering, accurately matching real data in spatial statistics and grain-size distribution [2211.10949].
- **Speech enhancement**: Unsupervised STFT-domain diffusion with EM posterior sampling yields competitive speech denoising, generalizing robustly to mismatched or unseen noise distributions [2309.10450].

Modern architectures typically leverage U-Net (or GNN for graphs) backbones with step/noise-level conditioning, self-attention, and classifier-free guidance for conditional generation [2311.08667, 2404.09016]. Convex optimization formulations for shallow network DSM objective yield exact solutions and non-asymptotic convergence in the case of two-layer ReLU networks [2402.01965].

## 6. Generalizations, Limitations, and Future Directions

EDMs admit broad generalizations:
- **Energy-based parameterizations and compositional energy priors**: Explicit modeling of energy enables simultaneous incorporation of multiple constraints via MH correction, with applications in compositional generation across vision, audio, and scientific computing [2510.22230].
- **Flexible SDE parameterizations**: Learning the spatial part of the forward SDE (e.g., Riemannian metric, Hamiltonian twist) extends the family of EDMs beyond fixed VP/VE SDEs, leading to unified and potentially better-optimized models [2206.10365].
- **Bridge processes and constrained-domain extensions**: Conditioning diffusion processes on endpoint or constraint sets enables direct modeling of discrete, categorical, and manifold data [2208.14699, 2302.02591].
- **Accelerated and equilibrium sampling**: Parallelized fixed-point solvers, high-order ODE/implicit samplers, or knowledge-distilled few-step students have addressed sampling speed bottlenecks, particularly in real-time or embedded contexts [2210.12867, 2311.08667].
- **Discrete data and non-Gaussian noise**: Categorical or non-Gaussian corruption models extend EDMs to sequences, graphs, and structured data [2404.09016, 2302.02591].

Limitations include high sampling cost (mitigated by accelerated samplers), lack of one-size-fits-all corruption kernels for general discrete data, difficulties in likelihood evaluation on structured data, and open questions on the tightness of theoretical bounds (especially in high dimensions). Evaluation metrics can inadvertently miss memorization or mode collapse in high-capacity models [2404.09016, 2311.08667].

Open theoretical directions focus on tight ELBO–NLL gaps, generalization in data-sparse regimes, and unification of statistical and algorithmic error analysis. Applications continue to broaden, including scientific inverse problems, classifier augmentation in imbalanced data, and compositional tasks requiring multi-energy integration [2510.22230, 2412.10948].

---

**Summary Table: Key Algorithmic and Theoretical Properties**

| Aspect                       | EDMs (classic/energy-based)         | Recent Theoretical Insights         |
|------------------------------|-------------------------------------|------------------------------------|
| Forward Mapping              | Gaussian Markov chain/SDE           | Abstract SDE, flexible metrics     |
| Reverse Process              | NN-parametrized score or mean       | ODE/SDE, energy-function, bridge   |
| Training Objective           | ELBO, DSM, Fisher divergence        | Convexification, finite-sample guarantees |
| Sampling Algorithm           | Markov, ODE (DDIM/DPM-Solver), MH-corrected | Accelerated non-asymptotic TV bounds |
| Domain Adaptation            | Images, audio, graphs, scientific   | Discrete, constrained, compositional |
| Notable Results              | SOTA generation, inverse sampling, microstructure reconstruction | $O(1/T)$/$O(1/T^2)$ TV rates, EM-style bridges |

For comprehensive reviews and technical depth, see [2510.22230], [2306.09251], [2404.09016], [2106.02808], [2311.08667], and [2412.10948].

Source: https://www.emergentmind.com/topics/diffusion-based-generative-models-edms