1. |
- Enqvist, Olof, et al.
(author)
-
A Brute-Force Algorithm for Reconstructing a Scene from Two Projections
- 2011
-
In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2011. - 1063-6919. ; , s. 2961-2968
-
Conference paper (peer-reviewed)abstract
- Is the real problem in finding the relative orientation of two viewpoints the correspondence problem? We argue that this is only one difficulty. Even with known correspondences, popular methods like the eight point algorithm and minimal solvers may break down due to planar scenes or small relative motions. In this paper, we derive a simple, brute-force algorithm which is both robust to outliers and has no such algorithmic degeneracies. Several cost functions are explored including maximizing the consensus set and robust norms like truncated least-squares. Our method is based on parameter search in a four-dimensional space using a new epipolar parametrization. In principle, we do an exhaustive search of parameter space, but the computations are very simple and easily parallelizable, resulting in an efficient method. Further speedups can be obtained by restricting the domain of possible motions to, for example, planar motions or small rotations. Experimental results are given for a variety of scenarios including scenes with a large portion of outliers. Further, we apply our algorithm to 3D motion segmentation where we outperform state-of-the-art on the well-known Hopkins-155 benchmark database.
|
|
2. |
- Jiang, Fangyuan, et al.
(author)
-
A Combinatorial Approach to L1-Matrix Factorization
- 2015
-
In: Journal of Mathematical Imaging and Vision. - : Springer Science and Business Media LLC. - 1573-7683 .- 0924-9907. ; 51:3, s. 430-441
-
Journal article (peer-reviewed)abstract
- Recent work on low-rank matrix factorization has focused on the missing data problem and robustness to outliers and therefore the problem has often been studied under the $L_1$-norm. However, due to the non-convexity of the problem, most algorithms are sensitive to initialization and tend to get stuck in a local optimum. In this paper, we present a new theoretical framework aimed at achieving optimal solutions to the factorization problem. We define a set of stationary points to the problem that will normally contain the optimal solution. It may be too time-consuming to check all these points, but we demonstrate on several practical applications that even by just computing a random subset of these stationary points, one can achieve significantly better results than current state of the art. In fact, in our experimental results we empirically observe that our competitors rarely find the optimal solution and that our approach is less sensitive to the existence of multiple local minima.
|
|
3. |
- Jiang, Fangyuan, et al.
(author)
-
A Minimal Solution to Relative Pose with Unknown Focal Length and Radial Distortion
- 2015
-
In: [Host publication title missing]. - Cham : Springer International Publishing. - 1611-3349 .- 0302-9743. ; 9004, s. 443-456
-
Conference paper (peer-reviewed)abstract
- In this paper, we study the minimal problem of estimating the essential matrix between two cameras with constant but unknown focal length and radial distortion. This problem is of both theoretical and practical interest and it has not been solved previously. We have derived a fast and stable polynomial solver based on Gr{\"o}bner basis method. This solver enables simultaneous auto-calibration of focal length and radial distortion for cameras. For experiments, the numerical stability of the solver is demonstrated on synthetic data. We also evaluate on real images using either RANSAC or kernel voting. Compared with the standard minimal solver, which does not model the radial distortion, our proposed solver both finds a larger set of geometrically correct correspondences on distorted images and gives an accurate estimate of the radial distortion and focal length.
|
|
4. |
- Jiang, Fangyuan, et al.
(author)
-
Association between antibiotic use during early life and early-onset colorectal cancer risk overall and according to polygenic risk and FUT2 genotypes
- 2023
-
In: International Journal of Cancer. - : John Wiley & Sons. - 0020-7136 .- 1097-0215. ; 153:9, s. 1602-1611
-
Journal article (peer-reviewed)abstract
- Early-onset colorectal cancer (EOCRC) has been increasing worldwide. Potential risk factors may have occurred in childhood or adolescence. We investigated the associations between early-life factors and EOCRC risk, with a particular focus on long-term or recurrent antibiotic use (LRAU) and its interaction with genetic factors. Data on the UK Biobank participants recruited between 2006 and 2010 and followed up to February 2022 were used. We used logistic regression to estimate adjusted odds ratios (ORs) and 95% confidence intervals (95% CIs) of the associations between LRAU during early life and EOCRC risk overall and by polygenic risk score (constructed by 127 CRC-related genetic variants) and Fucosyltransferase 2 (FUT2), a gut microbiota regulatory gene. We also assessed the associations for early-onset colorectal adenomas, as precursor lesion of CRC, to examine the effect of LRAU during early-life and genetic factors on colorectal carcinogenesis. A total of 113 256 participants were included in the analysis, with 165 EOCRC cases and 719 EOCRA cases. LRAU was nominally associated with increased risk of early-onset CRC (OR = 1.48, 95% CI = 1.01-2.17, P = .046) and adenomas (OR = 1.40, 95% CI = 1.17-1.68, P < .001). When stratified by genetic polymorphisms of FUT2, LRAU appeared to confer a comparatively greater risk for early-onset adenomas among participants with rs281377 TT genotype (OR = 1.10, 95% CI = 0.79-1.52, P = .587, for CC genotype; OR = 1.75, 95% CI = 1.16-2.64, P = .008, for TT genotype; Pinteraction = .089). Our study suggested that LRAU during early life is associated with increased risk of early-onset CRC and adenomas, and the association for adenomas is predominant among individuals with rs281377 TT/CT genotype. Further studies investigating how LRAU contributes together with genetic factors to modify EOCRC risk, particularly concerning the microbiome-related pathway underlying colorectal carcinogenesis, are warranted.
|
|
5. |
- Jiang, Fangyuan, et al.
(author)
-
Improved Object Detection and Pose Using Part-Based Models
- 2013
-
In: Lecture Notes in Computer Science (Image Analysis : 18th Scandinavian Conference, SCIA 2013, Espoo, Finland, June 17-20, 2013. Proceedings). - Berlin, Heidelberg : Springer Berlin Heidelberg. - 1611-3349 .- 0302-9743. - 9783642388859 - 9783642388866 ; 7944, s. 396-407
-
Conference paper (peer-reviewed)abstract
- Automated object detection is perhaps the most central task of computer vision and arguably the most difficult one. This paper extends previous work on part-based models by using accurate geometric models both in the learning phase and at detection. In the learning phase manual annotations are used to reduce perspective distortion before learning the part-based models. That training is performed on rectified images, leads to models which are more specific, reducing the risk of false positives. At the same time a set of representative object poses are learnt. These are used at detection to remove perspective distortion. The method is evaluated on the bus category of the Pascal dataset with promising results.
|
|
6. |
- Jiang, Fangyuan
(author)
-
Low Rank Matrix Factorization and Relative Pose Problems in Computer Vision
- 2015
-
Doctoral thesis (other academic/artistic)abstract
- This thesis is focused on geometric computer vision problems. The first part of the thesis aims at solving one fundamental problem, namely low-rank matrix factorization. We provide several novel insights into the problem. In brief, we characterize, generate, parametrize and solve the minimal problems associated with low-rank matrix factorization. Beyond that, we give several new algorithms based on the minimal solvers when the measurement matrix is either sparse, noisy or with outliers. The cost function and the algorithm can easily be adapted to several robust norms, for example, the L1-norm and the truncated L1-norm. We demonstrate our approach on several geometric computer vision problems. Another application is in sensor network calibration, which is also explored. The second part of the thesis deals with the relative pose problem. We solve the minimal problem of estimating the relative pose with unknown focal length and radial distortion. Beyond that, we also propose a brute force approach, which does not suffer from common algorithmic degeneracies. Further, the algorithm achieves a globally optimal solution up to a discretization error and it is easily parallelizable. Finally, we look into the problem of object detection with unknown pose.
|
|
7. |
- Jiang, Fangyuan, et al.
(author)
-
On the Minimal Problems of Low-Rank Matrix Factorization
- 2015
-
In: Computer Vision and Pattern Recognition (CVPR), 2015 IEEE Conference on. - 9781467369633 ; , s. 2549-2557
-
Conference paper (peer-reviewed)abstract
- Low-rank matrix factorization is an essential problem in many areas including computer vision, with applications in e.g. affine structure-from-motion, photometric stereo, and non-rigid structure from motion. However, very little attention has been drawn to minimal cases for this problem or to using the minimal configuration of observations to find the solution. Minimal problems are useful when either outliers are present or the observation matrix is sparse. In this paper, we first give some theoretical insights on how to generate all the minimal problems of a given size using Laman graph theory. We then propose a new parametrization and a building-block scheme to solve these minimal problems by extending the solution from a small sized minimal problem. We test our solvers on synthetic data as well as real data with outliers or a large portion of missing data and show that our method can handle the cases when other iterative methods, based on convex relaxation, fail.
|
|
8. |
- Jiang, Fangyuan, et al.
(author)
-
Relative pollen productivities of the major plant taxa of subtropical evergreen-deciduous mixed woodland in China
- 2020
-
In: Journal of Quaternary Science. - : John Wiley & Sons. - 0267-8179 .- 1099-1417. ; 35:4, s. 526-538
-
Journal article (peer-reviewed)abstract
- Quantifying the relationship between pollen and vegetation is an essential step in the pollen-based quantitative reconstruction of past vegetation cover. In this study, we use the Extended R-Value (ERV) model and a modern dataset of pollen (collected from moss polsters) and related vegetation from 50 sites in the Daba Mountains (subtropical China) to (i) estimate the relevant source area of pollen (RSAP) of the moss samples and the relative pollen productivities (RPPs) of nine major plant taxa-characteristic of the region, and (ii) evaluate the obtained RPPs. The RSAP estimates of moss polsters vary between 225 and 610 m depending on the ERV submodels and models of pollen dispersal and deposition used. The RPP estimates are different from values published in previous studies from temperate and subtropical China. This may be explained by differences in methodology, climate and vegetation (species composition and spatial distribution), of which vegetation is probably the most important factor. The ranking of the RPP estimates for the nine taxa is Pinus > Juglandaceae > D - Quercus (deciduous Quercus) > Poaceae > Rosaceae > Cyperaceae > Anacardiaceae > Castanea > Fabaceae. We use a 'leave-one-out' cross-validation strategy and the Landscape Reconstruction Algorithm (LRA) for pollen-based reconstruction of regional and local plant cover to evaluate the ERV model-based RPP estimates. Both the REVEALS (Regional Estimates of VEgetation Abundance from Large Sites)-based and the LOVE (LOcal Vegetation Estimates)-based plant cover using the RPP estimates are closer to the modern vegetation composition than pollen percentages, thus confirming the applicability of the ERV model and the LRA approach in subtropical China.
|
|
9. |
- Jiang, Fangyuan, et al.
(author)
-
Time Delay Estimation for TDOA Self-Calibration using Truncated Nuclear Norm Regularization
- 2013
-
In: Acoustics, Speech and Signal Processing (ICASSP), 2013 IEEE International Conference on. - 1520-6149. ; , s. 3885-3889
-
Conference paper (peer-reviewed)abstract
- Measurements with unknown time delays are common in different applications such as microphone array, radio antenna array calibration, where the sources (e.g. sounds) are transmitted in unknown time instants. In this paper, we present a method for estimating unknown time delays from Time-Difference-of-Arrival (TDOA) measurements. We propose a novel rank constraint on a matrix depending on the measurements and the unknown time delays. The time delays are recovered by solving a truncated nuclear norm minimization problem using alternating direction method of multipliers (ADMM). We show in synthetic experiments that the proposed method recovers the time delays with good accuracy for noisy and missing data.
|
|
10. |
- Larsson, Susanna C., et al.
(author)
-
Circulating lipoprotein(a) levels and health outcomes : Phenome-wide Mendelian randomization and disease-trajectory analyses
- 2022
-
In: Metabolism. - : Elsevier. - 0026-0495 .- 1532-8600. ; 137
-
Journal article (peer-reviewed)abstract
- BACKGROUND: Lipoprotein(a) [Lp(a)] is a risk factor for atherosclerotic and valvular diseases, but its possible role in other diseases has not yet been established. We conducted phenome-wide Mendelian randomization and disease-trajectory analyses to assess any associations of circulating Lp(a) levels with a broad range of diseases.METHODS: A weighted polygenic risk score was constructed using independent genetic variants in the LPA gene and with an established effect on Lp(a) levels. The PheWAS analysis included 1081 phenotype outcomes ascertained among 385,917 White participants of the UK Biobank. Novel findings were investigated in MR analysis using data from the FinnGen consortium. Disease-trajectory and comorbidity analyses were further conducted to explore the sequential patterns of multiple morbidities related to high circulating Lp(a) levels.RESULTS: PheWAS revealed statistically significant associations of higher circulating Lp(a) levels with increased risk of a large number of circulatory system diseases (including various cardiac diseases, peripheral vascular disease, hypertension, and valvular and cerebrovascular diseases) as well as some endocrine/metabolic diseases (including hyperlipidemia, hypercholesterolemia, disorders of lipoid metabolism, and type 2 diabetes), genitourinary system diseases (renal failure), and hematologic diseases (including different types of anemia). Two-sample MR analysis supported the association between Lp(a) and risk of anemia, showed a suggestive association with type 2 diabetes, but found no association with renal failure. Disease-trajectory and comorbidity analyses identified 3 major sequential patterns of multiple morbidities, mainly in the cardiovascular, metabolic, and mental disorders, related to high circulating Lp(a) levels.CONCLUSIONS: Genetically predicted higher circulating Lp(a) levels were associated with increased risk of many circulatory system diseases and anemia. Additionally, this study identified three major sequential patterns of multiple morbidities related to high Lp(a).
|
|