Filters








266 Hits in 7.1 sec

Scale Estimation of Monocular SfM for a Multi-modal Stereo Camera [article]

Shinya Sumikura, Ken Sakurada, Nobuo Kawaguchi, Ryosuke Nakamura
2018 arXiv   pre-print
This paper proposes a novel method of estimating the absolute scale of monocular SfM for a multi-modal stereo camera.  ...  In the fields of computer vision and robotics, scale estimation for monocular SfM has been widely investigated in order to simplify systems.  ...  Introduction This paper addresses the problem of estimating the scale parameter of monocular Structure from Motion (SfM) for a multi-modal stereo camera system (Fig. 1 ).  ... 
arXiv:1810.11856v1 fatcat:2bht26id6bbx5boeurumgea3fy

Two Stream Networks for Self-Supervised Ego-Motion Estimation [article]

Rares Ambrus, Vitor Guizilini, Jie Li, Sudeep Pillai, Adrien Gaidon
2019 arXiv   pre-print
Our experiments on a large-scale urban driving dataset of 1 million frames indicate that the performance of our proposed architecture does indeed scale progressively with more data.  ...  As a result, we show that our proposed two-stream pose network achieves state-of-the-art results among learning-based methods on the KITTI odometry benchmark, and is especially suited for self-supervision  ...  One of the earliest works in self-supervised depth estimation [8] used the photometric loss as a proxy for supervision to learn a monocular depth network from stereo imagery.  ... 
arXiv:1910.01764v2 fatcat:2njuchdotragzbf7mywbksundq

Dense Depth Estimation in Monocular Endoscopy with Self-supervised Learning Methods [article]

Xingtong Liu, Ayushi Sinha, Masaru Ishii, Gregory D. Hager, Austin Reiter, Russell H. Taylor, Mathias Unberath
2019 arXiv   pre-print
We present a self-supervised approach to training convolutional neural networks for dense depth estimation from monocular endoscopy data without a priori modeling of anatomy or shading.  ...  Our method only requires monocular endoscopic videos and a multi-view stereo method, e.g., structure from motion, to supervise learning in a sparse manner.  ...  Because of the inherent difficulty of global scale estimation of monocular camera-based methods, we elect to only estimate depth maps up to a global scale.  ... 
arXiv:1902.07766v2 fatcat:gzhawcczavah7przajrkfachj4

Recurrent Neural Network for Learning DenseDepth and Ego-Motion from Video [article]

Rui Wang, Jan-Michael Frahm, Stephen M. Pizer
2018 arXiv   pre-print
In this paper, we present a learning-based, multi-view dense depth map and ego-motion estimation method that uses Recurrent Neural Networks (RNN).  ...  There exists few learning-based, multi-view depth estimation methods.  ...  A particular case of dense geometry estimation is monocular depth estimation.  ... 
arXiv:1805.06558v1 fatcat:vwvstocrn5h2tkekek2w4afjpu

Endo-Depth-and-Motion: Reconstruction and Tracking in Endoscopic Videos using Depth Networks and Photometric Constraints [article]

David Recasens, José Lamarca, José M. Fácil, J. M. M. Montiel, Javier Civera
2021 arXiv   pre-print
In this paper we present Endo-Depth-and-Motion, a pipeline that estimates the 6-degrees-of-freedom camera pose and dense 3D scene models from monocular endoscopic videos.  ...  Estimating a scene reconstruction and the camera motion from in-body videos is challenging due to several factors, e.g. the deformation of in-body cavities or the lack of texture.  ...  INTRODUCTION Estimating a 3D reconstruction of a scene from a set of images, together with the poses of the cameras that captured them, is most of the times thought of as a mature technology.  ... 
arXiv:2103.16525v2 fatcat:3dmdzoxqx5c2xeto5pqetzzifa

Supervising the New with the Old: Learning SFM from SFM [chapter]

Maria Klodt, Andrea Vedaldi
2018 Lecture Notes in Computer Science  
Recent work has demonstrated that it is possible to learn deep neural networks for monocular depth and ego-motion estimation from unlabelled video sequences, an interesting theoretical development with  ...  We do so by using an off-the-shelf SFM system to generate a supervisory signal for the deep neural network.  ...  We are very grateful to Continental Corporation for sponsoring this research.  ... 
doi:10.1007/978-3-030-01249-6_43 fatcat:h4cpjsuucbg5bnukaugk4akyzi

Multimodal Scale Consistency and Awareness for Monocular Self-Supervised Depth Estimation [article]

Hemang Chawla, Arnav Varma, Elahe Arani, Bahram Zonooz
2021 arXiv   pre-print
The relative distance between frames captured through the GPS provides a scale signal that is independent of the camera setup and scene distribution, resulting in richer learned feature representations  ...  Dense depth estimation is essential to scene-understanding for autonomous driving.  ...  G represents the use of GPS for multi-modal selfsupervision).  ... 
arXiv:2103.02451v1 fatcat:uoa5ajjha5bnzdv5kefh44w7f4

Monocular Depth Estimation through Virtual-world Supervision and Real-world SfM Self-Supervision [article]

Akhil Gurram, Ahmet Faruk Tuna, Fengyi Shen, Onay Urfalioglu, Antonio M. López
2022 arXiv   pre-print
Usually, this GT is acquired at training time through a calibrated multi-modal suite of sensors. However, also using only a monocular system at training time is cheaper and more scalable.  ...  Nevertheless, problems of camouflaged objects, visibility changes, static-camera intervals, textureless areas, and scale ambiguity, diminish the usefulness of such self-supervision.  ...  Usually, such a GT is acquired at training time through a multi-modal suite of sensors, at least consisting of a camera calibrated with a LiDAR or some type of 3D laser scanner variant [5] [6] [7] [8]  ... 
arXiv:2103.12209v3 fatcat:4gejaemyqfbv7fcvdd5xqawaou

Indoor Visual SLAM Dataset With Various Acquisition Modalities

Imad El Bouazzaoui, Sergio Rodriguez, Bastien Vincke, Abdelhafid El Ouardi
2021 Data in Brief  
Each sequence is associated with a reference trajectory constructed with an Structure From Motion (SFM) and Multi View Stereo (MVS) library for comparison.  ...  The indoor Visual Simultaneous Localization And Mapping (V-SLAM) dataset with various acquisition modalities has been created to evaluate the impact of acquisition modalities on the Visual SLAM algorithm's  ...  A reference trajectory for comparison was created based on subsampled (subsampled by 1/5 for seq1 and seq2 and by 1/10 for seq3) monocular images of the environment using the SFM and MVS pipeline [7,  ... 
doi:10.1016/j.dib.2021.107496 pmid:34746344 pmcid:PMC8552193 fatcat:b44xmbxw2zbbziijkecsx3z5re

Deep Learning-Based Monocular Depth Estimation Methods—A State-of-the-Art Review

Faisal Khan, Saqib Salahuddin, Hossein Javidnia
2020 Sensors  
This survey provides a comprehensive overview of this research topic including the problem representation and a short description of traditional methods for depth estimation.  ...  Relevant datasets and 13 state-of-the-art deep learning-based approaches for monocular depth estimation are reviewed, evaluated and discussed.  ...  In the category of passive methods, there are two primary approaches: (a) multi-view depth estimation, such as depth from stereo, and (b) monocular depth estimation.  ... 
doi:10.3390/s20082272 pmid:32316336 pmcid:PMC7219073 fatcat:kfio24wembgxhkheafhd74gra4

Lidar-Monocular Surface Reconstruction Using Line Segments [article]

Victor Amblard, Timothy P. Osedach, Arnaud Croux, Andrew Speck, John J. Leonard
2021 arXiv   pre-print
One way to overcome this problem is to combine data from a monocular camera with that of a LIDAR.  ...  Structure from Motion (SfM) often fails to estimate accurate poses in environments that lack suitable visual features.  ...  ACKNOWLEDGEMENTS The authors wish to acknowledge Stéphane Vannuffelen, Sepand Ossia, and Kevin Doherty for insightful discussions and guidance.  ... 
arXiv:2104.02761v1 fatcat:h7pnsu7amvcyjcsyhxycgpuvzq

Multi-view Monocular Depth and Uncertainty Prediction with Deep SfM in Dynamic Environments [article]

Christian Homeyer, Oliver Lange, Christoph Schnörr
2022 arXiv   pre-print
3D reconstruction of depth and motion from monocular video in dynamic environments is a highly ill-posed problem due to scale ambiguities when projecting to the 2D image domain.  ...  In this work, we investigate the performance of the current State-of-the-Art (SotA) deep multi-view systems in such environments.  ...  As common in deep monocular multi-view SfM we report results for the ground truth scale aligned depth [72,69,70].  ... 
arXiv:2201.08633v1 fatcat:fnangw6thvafhpyuuycvy4ldiu

Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-shot Cross-dataset Transfer [article]

René Ranftl, Katrin Lasinger, David Hafner, Konrad Schindler, Vladlen Koltun
2020 arXiv   pre-print
Our approach clearly outperforms competing methods across diverse datasets, setting a new state of the art for monocular depth estimation.  ...  The success of monocular depth estimation relies on large and diverse training sets.  ...  It may be in the form of absolute depth (from laser-based measurements or stereo cameras with known calibration), depth up to an unknown scale (from SfM), or disparity maps (from stereo cameras with unknown  ... 
arXiv:1907.01341v3 fatcat:jzx54givkngcjhf7q3ruavwsni

Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-shot Cross-dataset Transfer

Rene Ranftl, Katrin Lasinger, David Hafner, Konrad Schindler, Vladlen Koltun
2020 IEEE Transactions on Pattern Analysis and Machine Intelligence  
Our approach clearly outperforms competing methods across diverse datasets, setting a new state of the art for monocular depth estimation.  ...  The success of monocular depth estimation relies on large and diverse training sets.  ...  It may be in the form of absolute depth (from laser-based measurements or stereo cameras with known calibration), depth up to an unknown scale (from SfM), or disparity maps (from stereo cameras with unknown  ... 
doi:10.1109/tpami.2020.3019967 pmid:32853149 fatcat:drtlguieu5elxmw2eshuwukfdy

Enhanced imaging colonoscopy facilitates dense motion-based 3D reconstruction

Pablo F. Alcantarilla, Adrien Bartoli, Francois Chadebecq, Christophe Tilmant, Vincent Lepilliez
2013 2013 35th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)  
We propose a novel approach for estimating a dense 3D model of neoplasia in colonoscopy using enhanced imaging endoscopy modalities.  ...  Then, the sparse reconstruction is densified using a Multi-View Stereo approach, and finally the dense 3D point cloud is transformed into a mesh by means of Poisson surface reconstruction.  ...  ACKNOWLEDGMENTS The authors would like to acknowledge the support of ANR through project SYSEO.  ... 
doi:10.1109/embc.2013.6611255 pmid:24111442 dblp:conf/embc/AlcantarillaBCT13 fatcat:of2k6rynd5dgxhmty3l5uek73y
« Previous Showing results 1 — 15 out of 266 results