AI Paper Atlas

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

Dataset Diffusion: Diffusion-based Synthetic Data Generation for Pixel-Level Semantic Segmentation

Quang Ho Nguyen, Truong Tuan Vu, Anh Tuan Tran, Khoi Nguyen

NeurIPS 2023 · Image generation › Text-to-image diffusion · approach: Compositional text-to-image guidance · 27 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
append class labels to captions (class-prompt appending) vs captions only (ablation)Dataset Diffusion · VOCmIoU: 50.8 -> 62.0Improved
append class labels to captions vs class labels only (ablation)Dataset Diffusion · VOCmIoU: 57.4 -> 62.0Improved
add self-attention exponentiation refinement to cross-attention (ablation)Dataset Diffusion · VOCmIoU: 44.8 -> 61.0Improved
add uncertainty-aware loss (ablation)Dataset Diffusion · VOCmIoU: 61.0 -> 62.0Improved
add self-training (ablation)Dataset Diffusion · VOCmIoU: 62.0 -> 62.7Improved
add test-time augmentation (ablation)Dataset Diffusion · VOCmIoU: 62.7 -> 64.3Improved
set self-attention exponent tau to 4 instead of 5 (ablation)Dataset Diffusion · VOCmIoU: 60.5 -> 62.0Improved
use uncertainty range 0.5-0.6 instead of 0.4-0.6 (ablation)Dataset Diffusion · VOCmIoU: 60.7 -> 62.0Improved
aggregate attention over all timesteps instead of subranges (ablation)Dataset Diffusion · VOCmIoU: 61.5-61.9 -> 62.0No clear effect
increase synthetic images from 10k to 40k (ablation)Dataset Diffusion · VOCmIoU: 63.8 -> 64.8Improved
use Dataset Diffusion synthetic data instead of DiffuMaskMask2Former R50 · VOC valmIoU: 57.4 -> 60.2Improved
add 20 labeled images to 5k synthetic dataDataset Diffusion · CelebA-HQ-MaskmIoU: 78.2 -> 89.9Improved

Later papers that compared with this method

5 later accepted papers report Dataset Diffusion as a baseline in their results tables; 5 of them report a better number on most of the results they share with it.

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

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