AI Paper Atlas

Research map / 3D vision / Gaussian splatting

GGRt: Towards Generalizable 3D Gaussians without Pose Priors in Real-Time

Hao Li, Yuanyuan Gao, Dingwen Zhang, Chenming Wu, YALUN DAI, Chen Zhao, Haocheng Feng, Errui Ding, Jingdong Wang, Junwei Han

ECCV 2024 · 3D vision › Gaussian splatting · approach: Compact and efficient Gaussian splatting

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 Gaussians cache (ablation)GGRt · Waymo 019inference time: 1.02s/iter -> 0.13s/iterImproved
add Gaussians cache (ablation)GGRt · Waymo 019PSNR: 32.34 -> 32.34No clear effect
add deferred back-propagation to enable 5 views at higher resolution (ablation)GGRt · LLFF roomPSNR: 26.00 -> 29.02 (generalized)Improved
train with cropped images instead of deferred back-propagation (ablation)GGRt · high-resolution renderingPSNR: 28.25 -> 24.22Hurt
add local self-attention in deferred back-propagation (ablation)GGRt · training PSNR curvetraining PSNRImproved
use joint training of IPO-Net with G-3DG instead of DBARF pose estimationpose-free NVS · LLFF rotation error (Trex)rotation error (degrees): 126.93 -> 3.449Improved
use GGRt instead of DBARFpose-free NVS · Waymo scene 69PSNR: 21.97 -> 32.10Improved
use GGRt (pose-free) instead of pose-based IBRNet on Waymo 126NVS · Waymo scene 126PSNR: 36.00 -> 34.15Hurt
rotation error on LLFF leaves with GGRt versus DBARFIPO-Net · LLFF leavesrotation error (degrees): 13.927 -> 16.964Hurt

Later papers that compared with this method

1 later accepted paper reports GGRT as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

Earlier methods this paper compares with

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