AI Paper Atlas

Research map / Image generation / Text-to-image diffusion

Your Diffusion Model is Secretly a Zero-Shot Classifier

Alexander C. Li, Mihir Prabhudesai, Shivam Duggal, Ellis Brown, Deepak Pathak

ICCV 2023 · Image generation › Text-to-image diffusion · approach: Compositional text-to-image guidance · 243 citations (OpenAlex)

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 changedWhereMeasuredEffect
use the diffusion model's noise-prediction error as a zero-shot classifier instead of training a ResNet-50 on synthetic SD imagesStable Diffusion 2.0 · ImageNet (zero-shot)accuracy: 18.9 -> 61.4Improved
use the same (t, eps) samples for all classes (difference testing) (ablation)Diffusion Classifier · Petsestimate accuracyImproved
sample timesteps evenly across [1,1000] instead of repeating a few intermediate timesteps (ablation)Diffusion Classifier · Petsaccuracy scalingImproved
use uniform l2 weighting instead of variational lower bound as objective (ablation)DiT-XL/2 Diffusion Classifier at 512^2 · ImageNetaccuracy: 74.0 -> 79.1Improved
use l1 error instead of squared l2 in zero-shot classifier (ablation)Stable Diffusion 2.0 · Zero-shot datasets (8)accuracy: ImageNet 61.4 -> 59.6; CIFAR10 84.4 -> 88.4Mixed
use l1 error instead of squared l2 in DiT supervised classifier (ablation)DiT-XL/2 at 512^2 · ImageNet-Aaccuracy: 30.2 -> 13.2Hurt
prune classes with CLIP ResNet-50 top-20 before diffusion scoring (ablation)Diffusion Classifier · ImageNet (zero-shot)accuracy / time per image: 61.4 -> 62.6 accuracy; 1000s -> 150sImproved
upgrade Stable Diffusion from 2.0 to 2.1 (ablation)Diffusion Classifier · ImageNet (zero-shot)accuracy: 61.4 -> 58.4Hurt
use later SD 1.x checkpoints with more training (ablation)Diffusion Classifier · ImageNet (zero-shot)accuracy: 51.7 -> 59.6 (SD 1.1 to 1.5)Improved
add classifier-free guidance to the noise-prediction error (ablation)Diffusion Classifier · zero-shot classificationaccuracyHurt
classify with generative Diffusion Classifier instead of discriminative classifierDiT-XL/2 vs ResNet/ViT · ImageNet-Aaccuracy: 20.8 -> 30.2 (ViT-B/16 vs 512^2)Improved
generative classifier effective robustness on ImageNetV2 and ObjectNetDiT-XL/2 Diffusion Classifier · ImageNetV2, ObjectNeteffective robustnessNo clear effect

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy