Are Large-Scale 3D Models Really Necessary for Accurate Visual Localization?

Sattler, Torsten; Torii, Akihiko; Šivic, Josef; Pollefeys, Marc; Taira, Hajime; Okutomi, Masatoshi; Pajdla, Tomáš

doi:10.1109/cvpr.2017.654

Cited by 176 publications

(174 citation statements)

References 57 publications

Supporting

Mentioning

171

Contrasting

Order By: Relevance

“…When performing localization in large-scale environments, matching a set of 2D keypoints with a large number of 3D landmarks can be difficult [60]. As suggested by [52], one way to reduce the set of 3D points to match the image keypoints against is to first perform image retrieval.…”

Section: Our Hierarchical Localization Pipelinementioning

confidence: 99%

Sparse-to-Dense Hypercolumn Matching for Long-Term Visual Localization

Germain

Bourmaud

Lepetit

2019

2019 International Conference on 3D Vision (3DV)

View full text Add to dashboard Cite

We propose a novel approach to feature point matching, suitable for robust and accurate outdoor visual localization in long-term scenarios. Given a query image, we first match it against a database of registered reference images, using recent retrieval techniques. This gives us a first estimate of the camera pose. To refine this estimate, like previous approaches, we match 2D points across the query image and the retrieved reference image. This step, however, is prone to fail as it is still very difficult to detect and match sparse feature points across images captured in potentially very different conditions. Our key contribution is to show that we need to extract sparse feature points only in the retrieved reference image: We then search for the corresponding 2D locations in the query image exhaustively. This search can be performed efficiently using convolutional operations, and robustly by using hypercolumn descriptors, i.e. image features computed for retrieval. We refer to this method as 'Sparse-to-Dense Hypercolumn Matching'. Because we know the 3D locations of the sparse feature points in the reference images thanks to an offline reconstruction stage, it is then possible to accurately estimate the camera pose from these matches. Our experiments show that this method allows us to outperform the state-of-the-art on several challenging outdoor datasets.

show abstract

Section: Our Hierarchical Localization Pipelinementioning

confidence: 99%

Sparse-to-Dense Hypercolumn Matching for Long-Term Visual Localization

Germain

Bourmaud

Lepetit

2019

2019 International Conference on 3D Vision (3DV)

View full text Add to dashboard Cite

show abstract

“…We apply the auxiliary camera pose on the visibility-wise match pool M d(k1) to realize the geometry-wise filtering. We define a relaxed reprojection error threshold θ in case rejecting potentially cor- [33] 20,862 6.77M 11,934 Aachen Day-Night [33] 4,328 1.65M 922 SF-0 [21,34] 610,773 30M 442 rect matches. As such, a match can be selected as a potentially correct match if the re-projection error with respect to the auxiliary camera pose is below θ.…”

Section: Geometry-wise Match Filteringmentioning

confidence: 99%

“…To evaluate under different levels of localization accuracy, we use the three accuracy intervals defined in [33] as follows: High-precision (0.25m, 2 • ), Medium-precision (0.5m, 5 • ) and Coarse-precision (5m, 10 • ). For the largescale SF-0 dataset [21], we use the evaluation package provided by [34] which contains reference camera poses for 442 query images.…”

Section: Datasets and Evaluation Metricsmentioning

confidence: 99%

Cascaded Parallel Filtering for Memory-Efficient Image-Based Localization

Cheng

Wang

Chen

et al. 2019

2019 IEEE/CVF International Conference on Computer Vision (ICCV)

View full text Add to dashboard Cite

Image-based localization (IBL) aims to estimate the 6DOF camera pose for a given query image. The camera pose can be computed from 2D-3D matches between a query image and Structure-from-Motion (SfM) models. Despite recent advances in IBL, it remains difficult to simultaneously resolve the memory consumption and match ambiguity problems of large SfM models. In this work, we propose a cascaded parallel filtering method that leverages the feature, visibility and geometry information to filter wrong matches under binary feature representation. The core idea is that we divide the challenging filtering task into two parallel tasks before deriving an auxiliary camera pose for final filtering. One task focuses on preserving potentially correct matches, while another focuses on obtaining high quality matches to facilitate subsequent more powerful filtering. Moreover, our proposed method improves the localization accuracy by introducing a quality-aware spatial reconfiguration method and a principal focal length enhanced pose estimation method. Experimental results on real-world datasets demonstrate that our method achieves very competitive localization performances in a memoryefficient manner.

show abstract

“…Research focuses on the reconstruction of such models and its automation [46][47][48]. Moreover, recently, various benchmark data sets for visual localization and with varying conditions have been published [6,49].…”

Section: Related Workmentioning

confidence: 99%

UAS Navigation with SqueezePoseNet—Accuracy Boosting for Pose Regression by Data Augmentation

Mueller

Jutzi

2018

Drones

View full text Add to dashboard Cite

Abstract:The navigation of Unmanned Aerial Vehicles (UAVs) nowadays is mostly based on Global Navigation Satellite Systems (GNSSs). Drawbacks of satellite-based navigation are failures caused by occlusions or multi-path interferences. Therefore, alternative methods have been developed in recent years. Visual navigation methods such as Visual Odometry (VO) or visual Simultaneous Localization and Mapping (SLAM) aid global navigation solutions by closing trajectory gaps or performing loop closures. However, if the trajectory estimation is interrupted or not available, a re-localization is mandatory. Furthermore, the latest research has shown promising results on pose regression in 6 Degrees of Freedom (DoF) based on Convolutional Neural Networks (CNNs). Additionally, existing navigation methods can benefit from these networks. In this article, a method for GNSS-free and fast image-based pose regression by utilizing a small Convolutional Neural Network is presented. Therefore, a small CNN (SqueezePoseNet) is utilized, transfer learning is applied and the network is tuned for pose regression. Furthermore, recent drawbacks are overcome by applying data augmentation on a training dataset utilizing simulated images. Experiments with small CNNs show promising results for GNSS-free and fast localization compared to larger networks. By training a CNN with an extended data set including simulated images, the accuracy on pose regression is improved up to 61.7% for position and up to 76.0% for rotation compared to training on a standard not-augmented data set.

show abstract

Are Large-Scale 3D Models Really Necessary for Accurate Visual Localization?

Cited by 176 publications

References 57 publications

Sparse-to-Dense Hypercolumn Matching for Long-Term Visual Localization

Sparse-to-Dense Hypercolumn Matching for Long-Term Visual Localization

Cascaded Parallel Filtering for Memory-Efficient Image-Based Localization

UAS Navigation with SqueezePoseNet—Accuracy Boosting for Pose Regression by Data Augmentation

Contact Info

Product

Resources

About