扩散 vs GAN 在质量、多样性、速度上的对比? 为什么文生图 SOTA 用扩散? (Compare diffusion vs GAN on quality, diversity and speed; why is diffusion the T2I SOTA?)

【扩散 vs GAN vs 自回归 (Diffusion vs GAN vs Autoregressive) 核心考点 1】扩散 vs GAN 在质量、多样性、速度上的对比? 为什么文生图 SOTA 用扩散?。权威解析：定义: 生成模型三大路线的对比, 核心是质量/多样性/速度三角。GAN 用生成器与判别器的零和博弈, 单步前向推理最快, 但对抗训练不稳定、易模式坍塌; 扩散/流模型通过逐步去噪生成, 训练是稳定回归、多样性好, 当前 SOTA FID 更低(质量上限更高); 自回归逐 token 预测, 擅长文本与离散模态、天然支持条件化, 但生成必须串行。对比表: | 维度 | GAN | 扩散 (DDPM/蒸馏) | 自回归 | |---|---|---|---| | 训练稳定性 | 差 (对抗不收敛) | 好 (MSE 回归) | 好 (交叉熵) | | 多样... 完整公式推导与落地权衡见：https://talentme.airsota.com/resources/tech/roadmap/mindmap/Multimodal/card/diffusion-vs-gan

写出 FID 公式并解释其含义, 为什么它不能完全代表多样性? (Write and explain the FID formula; why does it not fully capture diversity?)

【扩散 vs GAN vs 自回归 (Diffusion vs GAN vs Autoregressive) 核心考点 2】写出 FID 公式并解释其含义, 为什么它不能完全代表多样性?。权威解析：定义: 生成模型三大路线的对比, 核心是质量/多样性/速度三角。GAN 用生成器与判别器的零和博弈, 单步前向推理最快, 但对抗训练不稳定、易模式坍塌; 扩散/流模型通过逐步去噪生成, 训练是稳定回归、多样性好, 当前 SOTA FID 更低(质量上限更高); 自回归逐 token 预测, 擅长文本与离散模态、天然支持条件化, 但生成必须串行。对比表: | 维度 | GAN | 扩散 (DDPM/蒸馏) | 自回归 | |---|---|---|---| | 训练稳定性 | 差 (对抗不收敛) | 好 (MSE 回归) | 好 (交叉熵) | | 多样... 完整公式推导与落地权衡见：https://talentme.airsota.com/resources/tech/roadmap/mindmap/Multimodal/card/diffusion-vs-gan

👁️ MultimodalID: diffusion-vs-gan

Diffusion vs GAN vs Autoregressive

扩散 vs GAN vs 自回归

🎯Core Definition

The three major generation families trade off quality/diversity/speed. GANs play a zero-sum game between generator and discriminator — one forward pass is fastest, but adversarial training is unstable and prone to mode collapse. Diffusion/flow models generate by progressive denoising: training is stable regression and diversity is strong, and current SOTA FID is lower (higher quality ceiling). Autoregressive models predict token by token — great for text and discrete modalities with natural conditioning, but sampling is serial. Comparison:

📌Overview

| Dimension | GAN | Diffusion (DDPM/distilled) | Autoregressive | |---|---|---|---| | Training stability | poor (adversarial) | good (MSE regression) | good (cross-entropy) | | Diversity | poor (mode collapse) | good (full coverage) | good | | Quality (FID) | realistic in-mode | SOTA lowest FID | weaker on images | | Sampling speed | 1 step (fastest) | 4–1000 steps (DDIM/distillable) | serial tokens | | High-res scaling | unstable | latent + DiT scalable | hard |

📌Overview

FID (Fréchet Inception Distance): pass real and generated images through an Inception network, assume the features are Gaussian, and measure the Fréchet distance between their means and covariances:

📌Overview

FID = \Vert \mu_r - \mu_g \Vert^2 + \mathrm{Tr}\left(\Sigma_r + \Sigma_g - 2(\Sigma_r\Sigma_g)^{1/2}\right)

📌Overview

Lower is better; it captures both fidelity and diversity, though note they are not the same — a GAN can have decent FID yet poor diversity.

💡Use Cases

interview comparison questions (“why is text-to-image SOTA diffusion, not GAN?”); architecture selection — low-latency interactive tasks (real-time filters, SR) favor GANs or distilled diffusion, quality/diversity-first tasks (text-to-image, video) favor diffusion.

⚡Key Problems Solved

GANs' unstable adversarial training, mode collapse and hard high-resolution scaling; diffusion trades more sampling steps for stable training, full distribution coverage and lower FID; autoregressive models have been largely displaced for image generation by diffusion/flow models.

🎯Core Interview Exam Points

Fundamentals · Production Scenarios · Deep Dive

Core Def

Explain the core definition, underlying mechanism/math, and key principles of Diffusion vs GAN vs Autoregressive.

Use Cases

In production engineering, what are the primary use cases for Diffusion vs GAN vs Autoregressive and what core pain points does it solve?

🔥 High-Frequency Deep-Dive Questions

Compare diffusion vs GAN on quality, diversity and speed; why is diffusion the T2I SOTA?

Write and explain the FID formula; why does it not fully capture diversity?

Causes and detection of GAN mode collapse; how does diffusion naturally avoid it?

Pros and cons of autoregressive vs diffusion generation; why is AR being displaced on images?

For real-time use (mobile SR/filters), GAN or distilled diffusion — how to trade off latency and quality?

📖 In-depth Guide:📄 diffusion-models →

Updated 2026-08-12

🎯

Verify Mastery: Practice Exam Questions for "Diffusion vs GAN vs Autoregressive"

Take pitfall multiple-choice questions with derivation options; mistakes are automatically logged.

🚀 Start Card Practice ➔

Diffusion vs GAN vs Autoregressive

🔗 More Multimodal Knowledge Cards