AI Paper Atlas

Research map / Large language models / AI and society

CogCartoon: Towards Practical Story Visualization

Zhongyang Zhu, Jie Tang

IJCV 2024 · Large language models › AI and society · approach: Data commons and human baselines · 7 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 character-plugin generation and plugin-guided inference (CogCartoon) instead of Cones 2Stable Diffusion v2.1 · story 1 (girl)image alignment (IA): 0.695 -> 0.722Improved
use CogCartoon instead of Cones 2 on a three-character storyStable Diffusion v2.1 · story 7 (girl&boy&dog)image alignment (IA): 0.656 -> 0.659Improved
use CogCartoon instead of Cones 2 for text alignment on the boy storyStable Diffusion v2.1 · story 2 (boy)text alignment (TA): 0.319 -> 0.317Hurt
use CogCartoon instead of Cones 2 for text alignment on the boy&dog storyStable Diffusion v2.1 · story 6 (boy&dog)text alignment (TA): 0.340 -> 0.335Hurt
use CogCartoon instead of Cones 2 on human correspondence, single-character storyStable Diffusion v2.1 · story 1 (girl) human evalhuman correspondence (COR): 1.31 -> 2.87Improved
use CogCartoon instead of Cones 2 on human correspondence, three-character storyStable Diffusion v2.1 · story 7 (girl&boy&dog) human evalhuman correspondence (COR): 0.53 -> 1.73Improved
use CogCartoon instead of Cones 2 on human coherence, two-character storyStable Diffusion v2.1 · story 4 (girl&boy) human evalhuman coherence (COH): 0.23 -> 2.40Improved
use CogCartoon instead of Cones 2 on human quality, two-character storyStable Diffusion v2.1 · story 4 (girl&boy) human evalhuman quality (QUA): 0.10 -> 2.32Improved
add data augmentation (augmented images) to character-plugin training (ablation)CogCartoon · qualitative story generationvisual quality (qualitative)Improved
vary the number of augmented images (ablation)CogCartoon · qualitative story generationvisual quality (qualitative)No clear effect
use character-plugin instead of Residual Token Embedding (RTE) to represent characters (ablation)CogCartoon · qualitative story generationcharacter fidelity (qualitative)Improved

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