AI Paper Atlas

Research map / Vision-language and multimodal / Human and embodied modeling

Imagine the Unseen World: A Benchmark for Systematic Generalization in Visual World Models

Yeongbin Kim, Gautam Singh, Junyeong Park, Caglar Gulcehre, Sungjin Ahn

NeurIPS 2023 · Vision-language and multimodal › Human and embodied modeling · approach: Abstract visual reasoning benchmarks · 1 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 ViT encoder instead of CNN encoderImage-to-Image model · SVIB-dSprites EasyLPIPS: 0.0643 -> 0.0084Improved
use ViT encoder instead of CNN encoderImage-to-Image model · SVIB-dSprites MediumLPIPS: 0.1251 -> 0.0359Improved
use ViT encoder instead of CNN encoderImage-to-Image model · SVIB-CLEVR EasyLPIPS: 0.0593 -> 0.0583No clear effect
use ViT encoder instead of CNN encoderImage-to-Image model · SVIB-CLEVRTex EasyLPIPS: 0.2938 -> 0.2529Mixed
use slot-based latent (SSM-Slot) instead of single-vector VAE latentState-space model · SVIB-dSprites EasyLPIPS: 0.1284 -> 0.0534Improved
use slot-based latent (SSM-Slot) instead of single-vector VAE latentState-space model · SVIB-CLEVR HardLPIPS: 0.1884 -> 0.1139Improved
use slot-based latent (SSM-Slot) instead of single-vector VAE latentState-space model · SVIB-CLEVRTex HardLPIPS: 0.3745 -> 0.3590Improved
SSM-Slot instead of I2I-ViTbest state-space vs best image-to-image · SVIB-CLEVR MediumLPIPS: 0.1263 -> 0.0533Improved
SSM-Slot instead of I2I-ViTbest state-space vs best image-to-image · SVIB-dSprites HardLPIPS: 0.2326 -> 0.1834Improved
SSM-Slot instead of I2I-ViTbest state-space vs best image-to-image · SVIB-CLEVRTex EasyLPIPS: 0.2529 -> 0.3429Hurt
ground-truth tokens (Oracle) instead of learned perceptionOracle vs I2I-ViT · SVIB-CLEVRTex EasyLPIPS: 0.2529 -> 0.1414Improved
ground-truth tokens (Oracle) instead of learned perceptionOracle vs SSM-Slot · SVIB-dSprites HardLPIPS: 0.1834 -> 0.0064Improved

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