Surface Normal Estimation of Tilted Images via Spatial Rectifier

Do, Tien Van; Vuong, Khiem; Roumeliotis, Stergios I.; Park, Hyun Soo

doi:10.1007/978-3-030-58548-8_16

Cited by 21 publications

(11 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…III-B) on Azure Kinect datasets. Table III [5], and is trained with the truncated angular loss [39]. In these tests, we achieve satisfactory performance when evaluating DORN on the scenes with gravity aligned images (see Table III, Gravity-aligned frames).…”

Section: Generalization Capability On Azure Kinect Datasetsmentioning

confidence: 76%

“…To address this issue, we follow the approach of [39] to warp the input image so that the gravity, which is estimated from IMU, is aligned with the image's vertical axis. Note that, in this paper, we are interested in how the accuracy of surface normal prediction affects the depth completion performance, rather than solely focusing on the surface normal performance as in [39]. The idea of using gravity from VI-SLAM has been proposed in [40] as a regularization for a depth-prediction network during training time.…”

Section: B Surface Normal Networkmentioning

confidence: 99%

See 1 more Smart Citation

Deep Depth Estimation from Visual-Inertial SLAM

Sartipi¹,

Do²,

Ke³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

This paper addresses the problem of learning to complete a scene's depth from sparse depth points and images of indoor scenes. Specifically, we study the case in which the sparse depth is computed from a visual-inertial simultaneous localization and mapping (VI-SLAM) system. The resulting point cloud has low density, it is noisy, and has nonuniform spatial distribution, as compared to the input from active depth sensors, e.g., LiDAR or Kinect. Since the VI-SLAM produces point clouds only over textured areas, we compensate for the missing depth of the low-texture surfaces by leveraging their planar structures and their surface normals which is an important intermediate representation. The pretrained surface normal network, however, suffers from large performance degradation when there is a significant difference in the viewing direction (especially the roll angle) of the test image as compared to the trained ones. To address this limitation, we use the available gravity estimate from the VI-SLAM to warp the input image to the orientation prevailing in the training dataset. This results in a significant performance gain for the surface normal estimate, and thus the dense depth estimates. Finally, we show that our method outperforms other state-of-the-art approaches both on training (ScanNet [1] and NYUv2 [2]) and testing (collected with Azure Kinect [3]) datasets.

show abstract

Section: Generalization Capability On Azure Kinect Datasetsmentioning

confidence: 76%

Section: B Surface Normal Networkmentioning

confidence: 99%

Deep Depth Estimation from Visual-Inertial SLAM

Sartipi¹,

Do²,

Ke³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Surface Normal Prediction Despite having a continuous output domain like depth estimation, the surface normal problem remains under-explored for discretization approaches. Most prior works on surface normal estimation focus on improving the loss [4,25,58], reducing the distribution bias [4] and shift [25], and resolving the artifacts in ground-truth by leveraging other modalities [40,77]. Until recently, iDisc [75], a method based on discretizing internal representations, shows promise for depth estimation and 2), the probability distribution map is generated by first multiplying the per-pixel embeddings with the cluster centers, followed by a softmax operation.…”

Section: Related Workmentioning

confidence: 99%

Porosity and permeability prediction using a transformer and periodic long short-term network

et al. 2023

View full text Add to dashboard Cite

Effective reservoir parameter prediction is important for subsurface characterization and understanding the fluid migration. However, conventional methods for obtaining porosity and permeability are based on either core measurements or mathematical/petrophysical modeling, which are expensive or inefficient. In this study, we propose a reliable and low-cost deep learning (DL) framework for reservoir permeability and porosity prediction from real logging data at different regions. We leverage an advanced learning architecture (i.e., the transformer model) and design a new regression network (RPTransformer) that is sensitive to the depth period change of the logging data. The RPTransformer is composed of 1-D convolutional, long- and short-term memory (LSTM), and transformer layers. First, we use a 1-D convolutional layer for the first layer of the network to extract the significant features from logging data. Then, the nonlinear mapping relationships between logging data and reservoir parameters are established using several LSTM layers with a period parameter. Afterward, we use the encoder in the vision transformer (ViT) with the self-attention mechanism to further extract logging data features. The proposed network is a data-driven supervised learning framework and shows highly accurate and robust prediction results when applied to different geographic regions. To demonstrate the reliable prediction performance of our proposed network, we compare it with several classic machine learning and state-of-the-art DL methods, e.g., random forest (RF), multilayer LSTM (MLSTM), and LSTNet. More importantly, we show the generalization and uncertainty of the network in real-world applications through comprehensive numerical experiments.

show abstract

“…Normal vectors estimation [22][23][24] is the cornerstone of many 3D computer vision tasks such as segmentation [25], registration [26], surface construction [27], object recognition [28], and others. The most common approach to estimate the surface normal vector at a point is to fit a plane to a local neighborhood of the query point and determine the vector normal to the tangent plane (See Appendix A).…”

Section: Normal Estimation From Point Cloudsmentioning

confidence: 99%

Multiple Cylinder Extraction from Organized Point Clouds

Moradi

Laurendeau

Gosselin

2021

Sensors

View full text Add to dashboard Cite

Most man-made objects are composed of a few basic geometric primitives (GPs) such as spheres, cylinders, planes, ellipsoids, or cones. Thus, the object recognition problem can be considered as one of geometric primitives extraction. Among the different geometric primitives, cylinders are the most frequently used GPs in real-world scenes. Therefore, cylinder detection and extraction are of great importance in 3D computer vision. Despite the rapid progress of cylinder detection algorithms, there are still two open problems in this area. First, a robust strategy is needed for the initial sample selection component of the cylinder extraction module. Second, detecting multiple cylinders simultaneously has not yet been investigated in depth. In this paper, a robust solution is provided to address these problems. The proposed solution is divided into three sub-modules. The first sub-module is a fast and accurate normal vector estimation algorithm from raw depth images. With the estimation method, a closed-form solution is provided for computing the normal vector at each point. The second sub-module benefits from the maximally stable extremal regions (MSER) feature detector to simultaneously detect cylinders present in the scene. Finally, the detected cylinders are extracted using the proposed cylinder extraction algorithm. Quantitative and qualitative results show that the proposed algorithm outperforms the baseline algorithms in each of the following areas: normal estimation, cylinder detection, and cylinder extraction.

show abstract

Surface Normal Estimation of Tilted Images via Spatial Rectifier

Cited by 21 publications

References 33 publications

Deep Depth Estimation from Visual-Inertial SLAM

Deep Depth Estimation from Visual-Inertial SLAM

Porosity and permeability prediction using a transformer and periodic long short-term network

Multiple Cylinder Extraction from Organized Point Clouds

Contact Info

Product

Resources

About