AI Paper Atlas

Research map / Point clouds / Point cloud classification

AvatarPointillist: AutoRegressive 4D Gaussian Avatarization

Hongyu Liu, Xuan Wang, Zijian Wu, Yating Wang, Ziyu Wan, Yue Ma, Runtao Liu, Boyao Zhou, Yujun Shen, Qifeng Chen

CVPR 2026 · Point clouds › Point cloud classification · approach: Scene flow and dynamic point clouds · 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
Generate the Gaussian point cloud with the AR model instead of a fixed FLAME template (ablation)AvatarPointillist Gaussian decoder · NeRSemble self-reenactmentFID: 120.34 -> 95.18Improved
Generate the Gaussian point cloud with the AR model instead of a fixed FLAME template (ablation)AvatarPointillist Gaussian decoder · NeRSemble self-reenactmentLPIPS: 0.23 -> 0.15Improved
Add AR hidden features to the decoder input alongside positional embeddings (ablation)AvatarPointillist Gaussian decoder · NeRSemble self-reenactmentFID: 103.80 -> 95.18Improved
Add AR hidden features to the decoder input alongside positional embeddings (ablation)AvatarPointillist Gaussian decoder · NeRSemble self-reenactmentAKD: 5.81 -> 2.38Improved
Add positional encoding of generated points alongside AR features in the decoder input (ablation)AvatarPointillist Gaussian decoder · NeRSemble self-reenactmentFID: 110.93 -> 95.18Improved
Add positional encoding of generated points alongside AR features in the decoder input (ablation)AvatarPointillist Gaussian decoder · NeRSemble self-reenactmentAKD: 5.89 -> 2.38Improved
Use only AR features versus only positional embeddings as decoder input (ablation)AvatarPointillist Gaussian decoder · NeRSemble self-reenactmentAPD: 32.96 vs 41.49Mixed
Use AvatarPointillist instead of GAGAvatarsingle-image 4D avatar generation · NeRSemble cross-reenactmentAPD: 170.12 -> 153.13 (Portrait4Dv2 142.93 is lower)Hurt
Use AvatarPointillist instead of the best baseline on identity preservationsingle-image 4D avatar generation · NeRSemble cross-reenactmentCLIPScore: 0.71 -> 0.75Improved

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