AI Paper Atlas

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

L-Verse: Bidirectional Generation Between Image and Text

Taehoon Kim, Gwangmo Song, Sihaeng Lee, Sangyun Kim, Yewon Seo, Soonyoung Lee, Seung Hwan Kim, Honglak Lee, Kyunghoon Bae

CVPR 2022 · Image generation › Text-to-image diffusion · approach: Personalized and region-controlled generation · 25 citations (OpenAlex)

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 cross-level feature augmentation to the VQ-VAE codebookAugVAE · ImageNet1K validationFID: 3.28Improved
Use multi-level instead of single-level feature-augmented encoding (ablation)AugVAE · ImageNet1K validationFID: 3.28 -> 1.04Improved
Use 32-token truncated captions instead of full 64-token captions (ablation)L-Verse image captioning · MS-COCO Karpathy testCIDEr: 181.6 -> 102.2Hurt
Use 32-token truncated captions instead of full 64-token captions (ablation)L-Verse image captioning · MS-COCO Karpathy testSPICE: 28.9 -> 23.3Hurt
Use full 64-token captions instead of 32-token truncation (ablation)L-Verse image captioning · MS-COCO Karpathy testCIDEr: 102.2 -> 181.6Improved
Use full 64-token captions instead of 32-token truncation (ablation)L-Verse image captioning · MS-COCO Karpathy testSPICE: 23.3 -> 28.9Improved
Train on Conceptual Captions rather than MS-COCOL-Verse text-to-image · MS-COCO validation, 30,000 captionsFID-8: 29.8 -> 21.1Improved
Use L-Verse rather than prior MS-COCO-trained methodsText-to-image generation · MS-COCO validation, 30,000 captions, FID-2FID-2: 55.9 -> 35.5Improved
Use L-Verse rather than prior MS-COCO-trained methodsImage captioning · MS-COCO Karpathy testBLEU-4: 39.1 -> 39.9Improved
Use L-Verse rather than prior MS-COCO-trained methodsImage captioning · MS-COCO Karpathy testCIDEr: 131.2 -> 102.2Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 2 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