AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

Hierarchical Image Generation via Transformer-Based Sequential Patch Selection

Xiaogang Xu, Ning Xu

AAAI 2022 · Image generation › Autoregressive and latent generation · approach: Compositional scene generation · 10 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 proposed semi-parametric image generation frameworkScene-graph-to-image generation · COCO-Stuff, predicted boxesFID: 56.9 -> 51.6Improved
Use the proposed semi-parametric image generation frameworkScene-graph-to-image generation · Visual Genome, predicted boxesFID: 107.0 -> 63.7Improved
Use the proposed semi-parametric image generation frameworkScene-graph-to-image generation · COCO-Stuff, ground-truth boxesFID: 48.6 -> 46.9Improved
Use the proposed semi-parametric image generation frameworkScene-graph-to-image generation · Visual Genome, ground-truth boxesFID: 57.6 -> 56.7Improved
Replace PasteGAN retrieval with SCSM (ablation)PSGIM image generation · COCO-StuffFID: 58.9 -> 51.6Improved
Replace PasteGAN retrieval with SCSM (ablation)PSGIM image generation · Visual GenomeFID: 83.8 -> 63.7Improved
Replace RetrieveGAN retrieval with SCSM (ablation)PSGIM image generation · COCO-StuffFID: 55.3 -> 51.6Improved
Replace RetrieveGAN retrieval with SCSM (ablation)PSGIM image generation · Visual GenomeFID: 80.2 -> 63.7Improved
Add bounding-box position encoding to SCSM (ablation)SCSM + PSGIM · COCO-StuffFID: 54.7 -> 51.6Improved
Add bounding-box position encoding to SCSM (ablation)SCSM + PSGIM · Visual GenomeFID: 81.5 -> 63.7Improved
Use PSGIM instead of the PasteGAN generator (ablation)SCSM-based generation · COCO-StuffFID: 53.5 -> 51.6Improved
Use PSGIM instead of the PasteGAN generator (ablation)SCSM-based generation · Visual GenomeFID: 98.7 -> 63.7Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.

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