AI Paper Atlas

Research map / Video / Video generation

ShapeGen4D: Towards High Quality 4D Shape Generation from Videos

Jiraphon Yenphraphai, Ashkan Mirzaei, Jianqi Chen, Jiaxu Zou, Sergey Tulyakov, Raymond A. Yeh, Peter Wonka, Chaoyang Wang

ICLR 2026 · Video › Video generation · approach: Autoregressive long video generation · 11 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 aligned latents (ablation)ShapeGen4D · Objaverse held-out ablation (8 frames)Chamfer distance: 0.1348 -> 0.1096Improved
add shared noise across frames (ablation)ShapeGen4D · Objaverse held-out ablation (8 frames)F-Score: 0.2962 -> 0.3190Improved
add shared noise across frames (ablation)ShapeGen4D · Objaverse held-out ablation (8 frames)Chamfer distance: 0.1186 -> 0.1096Hurt
use full spatiotemporal attention instead of 1D temporal attention (ablation)ShapeGen4D · Objaverse held-out ablation (8 frames)Chamfer distance: 0.2118 -> 0.1096Improved
add image hidden states in spatiotemporal attention (ablation)ShapeGen4D · Objaverse held-out ablation (8 frames)F-Score: 0.3084 -> 0.3190Improved
add denoise time shift (ablation)ShapeGen4D · Objaverse held-out ablation (8 frames)Chamfer distance: 0.1374 -> 0.1096Improved
add shared noise to per-frame Step1X-3DStep1X-3D · Objaverse held-out test (33 samples)IoU: 0.3033 -> 0.3149Improved
add shared noise to per-frame Step1X-3DStep1X-3D · Objaverse held-out test (33 samples)Chamfer distance: 0.1356 -> 0.1368No clear effect
use ShapeGen4D instead of V2M4 (optimization-based)V2M4 · Objaverse held-out test (33 samples)Chamfer distance: 0.1233 -> 0.1220Improved
use ShapeGen4D instead of GVFDGVFD · Objaverse held-out test (33 samples)Chamfer distance: 0.3978 -> 0.1220Improved
use L4GM instead of ShapeGen4D for renderingShapeGen4D · Consistent4D (20 videos)LPIPS: 0.1359 -> 0.0988Hurt
use ShapeGen4D instead of Step1X-3D per frameStep1X-3D · Consistent4D (20 videos)FVD: 940 -> 796Improved

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