SOTAVerified

Camera Pose Estimation

Camera pose estimation is a crucial task in computer vision and robotics that involves determining the position and orientation (pose) of a camera relative to a given reference frame. This task is essential for various applications, such as augmented reality, 3D reconstruction, SLAM, and autonomous navigation.

Papers

Showing 125 of 304 papers

TitleStatusHype
Fast3R: Towards 3D Reconstruction of 1000+ Images in One Forward PassCode5
MonST3R: A Simple Approach for Estimating Geometry in the Presence of MotionCode5
SplaTAM: Splat, Track & Map 3D Gaussians for Dense RGB-D SLAMCode4
MV-DUSt3R+: Single-Stage Scene Reconstruction from Sparse Views In 2 SecondsCode4
One Diffusion to Generate Them AllCode4
Easi3R: Estimating Disentangled Motion from DUSt3R Without TrainingCode4
GIM: Learning Generalizable Image Matcher From Internet VideosCode4
SpatialTrackerV2: 3D Point Tracking Made EasyCode4
Cameras as Rays: Pose Estimation via Ray DiffusionCode4
LightGlue: Local Feature Matching at Light SpeedCode4
Uni4D: Unifying Visual Foundation Models for 4D Modeling from a Single VideoCode3
WHAC: World-grounded Humans and CamerasCode3
RoMa: Robust Dense Feature MatchingCode3
SGS-SLAM: Semantic Gaussian Splatting For Neural Dense SLAMCode3
PVO: Panoptic Visual OdometryCode2
DreamCar: Leveraging Car-specific Prior for in-the-wild 3D Car ReconstructionCode2
Recollection from Pensieve: Novel View Synthesis via Learning from Uncalibrated VideosCode2
ParticleSfM: Exploiting Dense Point Trajectories for Localizing Moving Cameras in the WildCode2
PoseDiffusion: Solving Pose Estimation via Diffusion-aided Bundle AdjustmentCode2
Reconstructing People, Places, and CamerasCode2
Learning to Produce Semi-dense Correspondences for Visual LocalizationCode2
CroCo: Self-Supervised Pre-training for 3D Vision Tasks by Cross-View CompletionCode2
HumanMM: Global Human Motion Recovery from Multi-shot VideosCode2
COLMAP-Free 3D Gaussian SplattingCode2
Investigating the Role of Image Retrieval for Visual Localization -- An exhaustive benchmarkCode2
Show:102550
← PrevPage 1 of 13Next →

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1Monodepth2Average Translational Error et[%]43.21Unverified
2SfMLearnerAverage Translational Error et[%]29.78Unverified
3GeoNetAverage Translational Error et[%]26.31Unverified
4SC-DepthAverage Translational Error et[%]12.2Unverified
5DeepMatchVOAverage Translational Error et[%]11.05Unverified
6SCIPaDAverage Translational Error et[%]8.63Unverified
7Manydepth2Average Translational Error et[%]7.15Unverified