Minimal Scene Descriptions from Structure from Motion Models

Cao, Song; Snavely, Noah

doi:10.1109/cvpr.2014.66

Cited by 85 publications

(97 citation statements)

References 22 publications

(41 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Image-Based Localization. Recent progress in imagebased localization has led to methods that are now quite robust to changes in scene appearance and illumination [4,61], scale to large scenes [43,56,58,83], and are suitable for realtime computation and mobile devices [8, 33, 35, 43-45, 57, 76] with compressed map representations [15,21]. Traditional localization methods based on image retrieval [34,66] and based on learning [12,35,80,81] have the advantage of not requiring the explicit storage of 3D maps.…”

Section: Related Workmentioning

confidence: 99%

Privacy Preserving Image-Based Localization

Speciale

Schönberger

Kang

et al. 2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

View full text Add to dashboard Cite

a) 3D Point Cloud (b) Inversion Attack on SIFT of 3D Point Cloud (c) 3D Line Cloud Projected 3D Points Reconstructed Image Original Image Figure 1: (a) Traditional image-based localization using 3D point cloud, which reveals potentially confidential information in the scene. (b) Reconstructed image using projected sparse 3D points and their SIFT features [50]. (c) Our proposed 3D line cloud protects user privacy by concealing the scene geometry and preventing inversion attacks, while still enabling accurate and efficient localization. AbstractImage-based localization is a core component of many augmented/mixed reality (AR/MR) and autonomous robotic systems. Current localization systems rely on the persistent storage of 3D point clouds of the scene to enable camera pose estimation, but such data reveals potentially sensitive scene information. This gives rise to significant privacy risks, especially as for many applications 3D mapping is a background process that the user might not be fully aware of. We pose the following question: How can we avoid disclosing confidential information about the captured 3D scene, and yet allow reliable camera pose estimation? This paper proposes the first solution to what we call privacy preserving image-based localization. The key idea of our approach is to lift the map representation from a 3D point cloud to a 3D line cloud. This novel representation obfuscates the underlying scene geometry while providing sufficient geometric constraints to enable robust and accurate 6-DOF camera pose estimation. Extensive experiments on several datasets and localization scenarios underline the high practical relevance of our proposed approach.

show abstract

Section: Related Workmentioning

confidence: 99%

Privacy Preserving Image-Based Localization

Speciale

Schönberger

Kang

et al. 2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

View full text Add to dashboard Cite

show abstract

“…A threshold of 224 was experimentally obtained from the model by evaluating corresponding descriptors of 3D points (similar to [5]), such that 95% of all correct matches survive. (D) A variable radius search, where the search radius is defined by 0.7 times the distance to the nearest neighbor in the query image itself.…”

Section: Experiments and Resultsmentioning

confidence: 99%

Camera Pose Voting for Large-Scale Image-Based Localization

Zeisl

Sattler

Pollefeys

2015

2015 IEEE International Conference on Computer Vision (ICCV)

166

207

View full text Add to dashboard Cite

Image-based localization approaches aim to determine the camera pose from which an image was taken. Finding correct 2D-3D correspondences between query image features and 3D points in the scene model becomes harder as the size of the model increases. Current state-of-theart methods therefore combine elaborate matching schemes with camera pose estimation techniques that are able to handle large fractions of wrong matches. In this work we study the benefits and limitations of spatial verification compared to appearance-based filtering. We propose a voting-based pose estimation strategy that exhibits O(n) complexity in the number of matches and thus facilitates to consider much more matches than previous approaches -whose complexity grows at least quadratically. This new outlier rejection formulation enables us to evaluate pose estimation for 1-to-many matches and to surpass the state-ofthe-art. At the same time, we show that using more matches does not automatically lead to a better performance.

show abstract

“…Impact of using 2D-2D point correspondences Results for networks trained without the additional dataset from [42] or with the correspondence loss disabled (where the clustering still is done on feature from the CMU/RobotCar images), are shown in Table 2 (row [11][12][13][14].…”

Section: Visual Localizationmentioning

confidence: 99%

Fine-Grained Segmentation Networks: Self-Supervised Segmentation for Improved Long-Term Visual Localization

Larsson

Stenborg

Toft

et al. 2019

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

View full text Add to dashboard Cite

Long-term visual localization is the problem of estimating the camera pose of a given query image in a scene whose appearance changes over time. It is an important problem in practice, for example, encountered in autonomous driving. In order to gain robustness to such changes, long-term localization approaches often use segmantic segmentations as an invariant scene representation, as the semantic meaning of each scene part should not be affected by seasonal and other changes. However, these representations are typically not very discriminative due to the limited number of available classes. In this paper, we propose a new neural network, the Fine-Grained Segmentation Network (FGSN), that can be used to provide image segmentations with a larger number of labels and can be trained in a self-supervised fashion. In addition, we show how FGSNs can be trained to output consistent labels across seasonal changes. We demonstrate through extensive experiments that integrating the fine-grained segmentations produced by our FGSNs into existing localization algorithms leads to substantial improvements in localization performance.

show abstract

Minimal Scene Descriptions from Structure from Motion Models

Cited by 85 publications

References 22 publications

Privacy Preserving Image-Based Localization

Privacy Preserving Image-Based Localization

Camera Pose Voting for Large-Scale Image-Based Localization

Fine-Grained Segmentation Networks: Self-Supervised Segmentation for Improved Long-Term Visual Localization

Contact Info

Product

Resources

About