[ECCV 2024 - Oral] ACE0 is a learning-based structure-from-motion approach that estimates camera parameters of sets of images by learning a multi-view consistent, implicit scene representation.

[IROS 2024] Representing 3D sparse map points and lines for camera relocalization

DFNet: Enhance Absolute Pose Regression with Direct Feature Matching (ECCV 2022)

VidLoc: A Deep Spatio-Temporal Model for 6-DoF Video-Clip Relocalization, Geometric loss functions for camera pose regression with deep learning, PoseNet: A Convolutional Network for Real-Time 6-DOF Camera Relocalization

Hierarchical Scene Coordinate Classification and Regression for Visual Localization

(CVPR 2024) Neural Refinement for Absolute Pose Regression with Feature Synthesis

Direct-PoseNet: Absolute Pose Regression with Photometric Consistency (3DV 2021)

A CNN-based local image descriptor

A PyTorch implementation of "Camera Relocalization by Computing Pairwise Relative Poses Using Convolutional Neural Network" by Z.Laskar et al.