Eric Brachmann

I am a staff scientist at Niantic, working on the Lightship Visual Positioning System (VPS). I work at the intersection of machine learning and computer vision, 3D vision in particular. My research revolves around topics such as visual relocalisation, pose estimation, end-to-end learning, robust optimization and feature matching.

I publish my research in the top conferences in computer vision where I am also active as area chair and reviewer with several outstanding reviewer mentions. I co-organized several tutorials and workshops on visual relocalisation and object pose estimation.

E-Mail Google Scholar Twitter LinkedIn CV
Scene Coordinate Reconstruction: Posing of Image Collections via Incremental Learning of a RelocalizerECCV 2024 oral

Eric Brachmann, Jamie Wynn, Shuai Chen, Tommaso Cavallari, Áron Monszpart, Daniyar Turmukhambetov, Victor Adrian Prisacariu

TL;DR: self-supervised ACE = learning-based structure-from-motion, needs no pose priors, works on unordered image sets, efficiently handles thousands of images.

  arXiv project page   code   video
Matching 2D Images in 3D: Metric Relative Pose from Metric Correspondences CVPR 2024 oral

Axel Barroso-Laguna, Sowmya Munukutla, Victor Adrian Prisacariu, Eric Brachmann

TL;DR: MicKey, a method that regresses and matches scale-metric 3D key points, trained end-to-end using differentiable RANSAC

  arXiv project page   code
Accelerated Coordinate Encoding: Learning to Relocalize in Minutes using RGB and Poses CVPR 2023 highlight

Eric Brachmann, Tommaso Cavallari, Victor Adrian Prisacariu

TL;DR: creating maps in 5 minutes with SOTA accuracy, up to 300x faster mapping than DSAC*, maps are 4MB large, new dataset

  arXiv project page   blog   code   dataset   video
Map-Free Visual Relocalization: Metric Pose Relative to a Single Image ECCV 2022

Eduardo Arnold, Jamie Wynn, Sara Vicente, Guillermo Garcia-Hernando, Aron Monszpart, Victor Prisacariu, Daniyar Turmukhambetov, Eric Brachmann

TL;DR: only one mapping image and one query, dataset with multiple hundred outdoor scenes, benchmark and online leaderboard

arXiv supplement project page code dataset video
Visual Camera Re-localization From RGB and RGB-D Images using DSAC TPAMI 2021

Eric Brachmann, Carsten Rother

TL;DR: DSAC*, higher accuracy than DSAC++ and full depth support, 28MB standard maps, 4MB tiny maps

arXiv code
On the Limits of Pseudo Ground Truth in Visual Camera Re-localisation ICCV 2021

Eric Brachmann, Martin Humenberger, Carsten Rother, Torsten Sattler

TL;DR: the choice of algorithm to generate reference poses, SfM or D-SLAM, has large impact on the ranking or relocalizers

arXiv code video
Reinforced Feature Points: Optimizing Feature Detection and Description for a High-Level Task CVPR 2020 oral

Aritra Bhowmik, Stefan Gumhold, Carsten Rother, Eric Brachmann

TL;DR: refine SuperPoint end-to-end for relative pose estimation, gradients of feature matching wrt feature descriptors and key point heatmap

arXiv code video
Neural-Guided RANSAC: Learning Where to Sample Model HypothesisICCV 2019

Eric Brachmann, Carsten Rother

TL;DR: NG-RANSAC + NG-DSAC, gradients of RANSAC-fitted model wrt quality of data points, applied to E/F matrix fitting, horizon line estimation and camera relocalization

arXiv project page F/E matrix code horizon line code relocalisation code video
BOP: Benchmark for 6D Object Pose EstimationECCV 2018

Tomas Hodan, Frank Michel, Eric Brachmann, Wadim Kehl, Anders Buch, Dirk Kraft, Bertram Drost, Joel Vidal, Stephan Ihrke , Xenophon Zabulis, Caner Sahin, Fabian Manhardt, Federico Tombari, Tae-Kyun Kim, Jiri Matas, Carsten Rother

TL;DR: de facto standard benchmark for instance pose estimation, unifying dataset formats and proposing evaluation metrics, ongoing competition with online leaderboard

arXiv project page
Learning Less is More - 6D Camera Localization via 3D Surface Regression CVPR 2018

Eric Brachmann, Carsten Rother

TL;DR: DSAC++, first time training scene coordinate regression without depth, differentiable PnP

arXiv project page code video
DSAC - Differentiable RANSAC for Camera Localization CVPR 2017 oral

Eric Brachmann, Alexander Krull, Sebastian Nowozin, Jamie Shotton, Frank Michel, Stefan Gumhold, Carsten Rother

TL;DR: gradients of a RANSAC-fitted model wrt the coordinates of the input points, using policy gradient on discrete hypothesis selection

arXiv project page toy code relocalisation code video
Learning 6D Object Pose Estimation using 3D Object Coordinates ECCV 2014

Eric Brachmann, Alexander Krull, Frank Michel, Stefan Gumhold, Jamie Shotton, Carsten Rother

TL;DR: introduces dense image-to-object correspondences as a learnable intermediate representation, introduced the LINEMOD-Occlusion dataset

paper supplement project page dataset video 1 video 2
Scene Coordinate Reconstruction: Posing of Image Collections via Incremental Learning of a RelocalizerECCV 2024 oral

Eric Brachmann, Jamie Wynn, Shuai Chen, Tommaso Cavallari, Áron Monszpart, Daniyar Turmukhambetov, Victor Adrian Prisacariu

TL;DR: self-supervised ACE = learning-based structure-from-motion, needs no pose priors, works on unordered image sets, efficiently handles thousands of images.

  arXiv project page   code   video
Map-Relative Pose Regression for Visual Re-Localization CVPR 2024 highlight

Shuai Chen, Tommaso Cavallari, Victor Adrian Prisacariu, Eric Brachmann

TL;DR: marepo, a scene-agnostic absolute pose regression transformer on top of a scene-specific ACE map representation, on-par with structure-based relocalizers in terms of accuracy and mapping time

  arXiv project page   code
Matching 2D Images in 3D: Metric Relative Pose from Metric Correspondences CVPR 2024 oral

Axel Barroso-Laguna, Sowmya Munukutla, Victor Adrian Prisacariu, Eric Brachmann

TL;DR: MicKey, a method that regresses and matches scale-metric 3D key points, trained end-to-end using differentiable RANSAC

  arXiv project page   code
BOP Challenge 2023 on Detection, Segmentation and Pose Estimation of Seen and Unseen Rigid Objects CVPR Workshops 2024

Tomas Hodan, Martin Sundermeyer, Yann Labbe, Van Nguyen Nguyen, Gu Wang, Eric Brachmann, Bertram Drost, Vincent Lepetit, Carsten Rother, Jiri Matas

TL;DR: results of BOP challenge 2023, accuracy is excellent if objects are known in advance, for unseen objects, still good but slow

arXiv project page video
Robust Shape Fitting for 3D Scene Abstraction TPAMI 2024

Florian Kluger, Eric Brachmann, Michael Ying Yang, Bodo Rosenhahn

TL;DR: extended version of "Cuboids Revisited" (CVPR 2021), a neural solver fitting cuboids to 3D points leads to better scene abstractions and faster runtime

arXiv code
Accelerated Coordinate Encoding: Learning to Relocalize in Minutes using RGB and Poses CVPR 2023 highlight

Eric Brachmann, Tommaso Cavallari, Victor Adrian Prisacariu

TL;DR: creating maps in 5 minutes with SOTA accuracy, up to 300x faster mapping than DSAC*, maps are 4MB large, new dataset

  arXiv project page   blog   code   dataset   video
Two-View Geometry Scoring Without Correspondences CVPR 2023

Axel Barroso-Laguna, Eric Brachmann, Victor Adrian Prisacariu, Gabriel J. Brostow, Daniyar Turmukhambetov

TL;DR: inlier counting is unreliable for selecting pose hypotheses when correspondence count is low, instead train a transformer to score hypotheses

paper paper code
BOP Challenge 2022 on Detection, Segmentation and Pose Estimation of Specific Rigid Objects CVPR Workshops 2023

Martin Sundermeyer, Tomas Hodan, Yann Labbe, Gu Wang, Eric Brachmann, Bertram Drost, Carsten Rother, Jiri Matas

TL;DR: results of BOP challenge 2022, deep neural networks beat everything else

arXiv project page video 1 video 2
Map-Free Visual Relocalization: Metric Pose Relative to a Single Image ECCV 2022

Eduardo Arnold, Jamie Wynn, Sara Vicente, Guillermo Garcia-Hernando, Aron Monszpart, Victor Prisacariu, Daniyar Turmukhambetov, Eric Brachmann

TL;DR: only one mapping image and one query, dataset with multiple hundred outdoor scenes, benchmark and online leaderboard

arXiv supplement project page code dataset video
Camera Pose Estimation and Localization with Active Audio Sensing ECCV 2022

Karren Yang, Michael Firman, Eric Brachmann, Clement Godard

TL;DR: camera pose by echolocation, relative pose / absolute pose / image retrieval, vision is more accurate but sound helps when vision fails

paper
Visual Camera Re-localization From RGB and RGB-D Images using DSAC TPAMI 2021

Eric Brachmann, Carsten Rother

TL;DR: DSAC*, higher accuracy than DSAC++ and full depth support, 28MB standard maps, 4MB tiny maps

arXiv code
Visual Camera Re-localization using Graph Neural Networks and Relative Pose Supervision 3DV 2021 oral

Mehmet Ozgur Turkoglu, Eric Brachmann, Konrad Schindler, Gabriel Brostow, Aron Monszpart

TL;DR: relative pose regression, trained scene-agnostic, propagate information from kNN mapping images to query

arXiv code
On the Limits of Pseudo Ground Truth in Visual Camera Re-localisation ICCV 2021

Eric Brachmann, Martin Humenberger, Carsten Rother, Torsten Sattler

TL;DR: the choice of algorithm to generate reference poses, SfM or D-SLAM, has large impact on the ranking or relocalizers

arXiv code video
Cuboids Revisited: Learning Robust 3D Shape Fitting to Single RGB Images CVPR 2021

Florian Kluger, Hanno Ackermann, Eric Brachmann, Michael Ying Yang, Bodo Rosenhahn

TL;DR: sequentially fit cuboids to an estimated depth map, box representation of complex indoor scenes

arXiv code video
BOP Challenge 2020 on 6D Object Localization ECCV Workshops 2020

Tomas Hodan, Martin Sundermeyer, Bertram Drost, Yann Labbe, Eric Brachmann, Frank Michel, Carsten Rother, Jiri Matas

TL;DR: results of BOP challenge 2020, deep neural networks on par with point pair features

arXiv project page
Reinforced Feature Points: Optimizing Feature Detection and Description for a High-Level Task CVPR 2020 oral

Aritra Bhowmik, Stefan Gumhold, Carsten Rother, Eric Brachmann

TL;DR: refine SuperPoint end-to-end for relative pose estimation, gradients of feature matching wrt feature descriptors and key point heatmap

arXiv code video
CONSAC: Robust Multi-Model Fitting by Conditional Sample Consensus CVPR 2020

Florian Kluger, Eric Brachmann, Hanno Ackermann, Carsten Rother, Michael Yang, Bodo Rosenhahn

TL;DR: repeated application of NG-RANSAC to find parameters of N models, learned sequential search while updating internal state

arXiv code dataset 1 dataset 2 video
Expert Sample Consensus Applied to Camera Re-LocalizationICCV 2019

Eric Brachmann, Carsten Rother

TL;DR: ESAC, end-to-end learning of mixture-of-experts and RANSAC, large scale scene coordinate regression

arXiv project page code video
Neural-Guided RANSAC: Learning Where to Sample Model HypothesisICCV 2019

Eric Brachmann, Carsten Rother

TL;DR: NG-RANSAC + NG-DSAC, gradients of RANSAC-fitted model wrt quality of data points, applied to E/F matrix fitting, horizon line estimation and camera relocalization

arXiv project page F/E matrix code horizon line code relocalisation code video
iPose: Instance-Aware 6D Pose Estimation of Partly Occluded ObjectsACCV 2018

Omid Hosseini Jafari, Siva Karthik Mustikovela, Karl Pertsch, Eric Brachmann, Carsten Rother

TL;DR: instance segmentation + deep object coordinate prediction

arXiv
BOP: Benchmark for 6D Object Pose EstimationECCV 2018

Tomas Hodan, Frank Michel, Eric Brachmann, Wadim Kehl, Anders Buch, Dirk Kraft, Bertram Drost, Joel Vidal, Stephan Ihrke , Xenophon Zabulis, Caner Sahin, Fabian Manhardt, Federico Tombari, Tae-Kyun Kim, Jiri Matas, Carsten Rother

TL;DR: de facto standard benchmark for instance pose estimation, unifying dataset formats and proposing evaluation metrics, ongoing competition with online leaderboard

arXiv project page
Learning to Predict Dense Correspondences for 6D Pose Estimation PhD thesis

Eric Brachmann

TL;DR: summary of my work prior to 2018, learning object and scene coordinate regression using random forests and neural networks

thesis
Learning Less is More - 6D Camera Localization via 3D Surface Regression CVPR 2018

Eric Brachmann, Carsten Rother

TL;DR: DSAC++, first time training scene coordinate regression without depth, differentiable PnP

arXiv project page code video
DSAC - Differentiable RANSAC for Camera Localization CVPR 2017 oral

Eric Brachmann, Alexander Krull, Sebastian Nowozin, Jamie Shotton, Frank Michel, Stefan Gumhold, Carsten Rother

TL;DR: gradients of a RANSAC-fitted model wrt the coordinates of the input points, using policy gradient on discrete hypothesis selection

arXiv project page toy code relocalisation code video
Global Hypothesis Generation for 6D Object Pose Estimation CVPR 2017 spotlight

Frank Michel, Alexander Kirillov, Eric Brachmann, Alexander Krull, Stefan Gumhold, Bogdan Savchynskyy, Carsten Rother

TL;DR: find pose inlier correspondences by optimizing the energy in a graphical model

arXiv project page
PoseAgent: Budget-Constrained 6D Object Pose Estimation via Reinforcement Learning CVPR 2017

Alexander Krull, Eric Brachmann, Sebastian Nowozin, Frank Michel, Jamie Shotton, Carsten Rother

TL;DR: an RL agent chooses which RANSAC hypothesis to refine next

arXiv project page
Random Forests versus Neural Networks - What's Best for Camera Relocalization? ICRA 2017

Daniela Massiceti, Alexander Krull, Eric Brachmann, Carsten Rother, Philip H.S. Torr

TL;DR: mapping of random forests to NNs for optimization, and back again for efficiency

arXiv
Uncertainty-Driven 6D Pose Estimation of Objects and Scenes from a Single RGB Image CVPR 2016

Eric Brachmann, Frank Michel, Alexander Krull, Michael Ying Yang, Stefan Gumhold, Carsten Rother

TL;DR: first object/scene coordinate regression system for RGB, predict correspondence distributions and search for max likelihood pose

paper supplement project page video
Learning Analysis-by-Synthesis for 6D Pose Estimation in RGB-D Images ICCV 2015

Alexander Krull, Eric Brachmann, Frank Michel, Michael Ying Yang, Stefan Gumhold, Carsten Rother

TL;DR: substitute inlier counting pose score with a CNN that compares input image and renderings, trained via max likelihood

paper supplement project page video
Pose Estimation of Kinematic Chain Instances via Object Coordinate Regression BMVC 2015 oral

Frank Michel, Alexander Krull, Eric Brachmann, Michael Ying Yang, Stefan Gumhold, Carsten Rother

TL;DR: only n+2 correspondences are needed to estimate pose of n-jointed objects

paper conference page project page
6-DOF Model Based Tracking via Object Coordinate Regression ACCV 2014 oral Honorable Mention Demo Award

Alexander Krull, Frank Michel, Eric Brachmann, Stefan Gumhold, Stephan Ihrke, Carsten Rother

TL;DR: combines RANSAC-based hypothesis sampling with particle filter for real-time pose tracking

paper supplement project page video 1 video 2
Learning 6D Object Pose Estimation using 3D Object Coordinates ECCV 2014

Eric Brachmann, Alexander Krull, Frank Michel, Stefan Gumhold, Jamie Shotton, Carsten Rother

TL;DR: introduces dense image-to-object correspondences as a learnable intermediate representation, introduced the LINEMOD-Occlusion dataset

paper supplement project page dataset video 1 video 2
Feature Propagation on Image Webs for Enhanced Image Retrieval ICMR 2013 oral

Eric Brachmann, Marcel Spehr, Stefan Gumhold

TL;DR: propagate visual words along image web edges to make a BoW image descriptors more robust

paper
Simplified Authentication and Authorization for RESTful Services in Trusted Environments ESOCC 2012

Eric Brachmann, Gero Dittmann, Klaus-Dieter Schubert

TL;DR: an authentication scheme for company intranets where you may want to trade security for simplicity

paper

Feel free to use this website as template. Inspired by Jon Barron's iconic template.