AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

DatasetDM: Synthesizing Data with Perception Annotations Using Diffusion Models

Weijia Wu, Yuzhong Zhao, Hao Chen, Yuchao Gu, Rui Zhao, Yefei He, Hong Zhou, Mike Zheng Shou, Chunhua Shen

NeurIPS 2023 · Image generation › Fast diffusion sampling · approach: Diffusion models as perception tools · 32 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
add 40k DatasetDM synthetic images to 100 real imagesMask2Former Swin-B · VOC 2012mIoU: 65.2 -> 78.5Improved
add synthetic data to full real training setMask2Former Swin-B · VOC 2012 fullmIoU: 84.3 -> 85.4Improved
add synthetic data to 800 real imagesMask2Former Swin-B · COCO instance segAP: 14.4 -> 26.5Improved
add synthetic data to 50 real imagesDepthformer · NYU Depth V2REL: 0.31 -> 0.21Improved
synthetic long-tail data for tail classesMask2Former · VOC long-tailmIoU: 52.6 -> 70.0Improved
synthetic only vs DiffuMask on unseen classesDatasetDM · VOC zero-shotunseen mIoU: 65.0 -> 60.5Hurt
use larger diffusion step 800 instead of step 1 (ablation)DatasetDM · VOC 2012mIoU: 78.5 -> 76.1Hurt
add cross-attention maps at step 1 (ablation)DatasetDM · VOC 2012mIoU: 77.6 -> 78.5Improved
increase training images 60 -> 1,000 (ablation)DatasetDM · VOC 2012mIoU (joint): 77.1 -> 81.1Improved
GPT-4 prompts instead of human prompts (ablation)DatasetDM · VOC 2012mIoU: 76.6 -> 77.1Improved
increase GPT-4 prompt candidates 100 -> 1k (ablation)DatasetDM · VOC 2012mIoU: 77.1 -> 78.5Improved
increase synthetic images 100 -> 40k (ablation)DatasetDM joint · VOC 2012mIoU: 67.9 -> 78.5Improved

Later papers that compared with this method

3 later accepted papers report DatasetDM as a baseline in their results tables; 3 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