Robust and Accurate Shape Model Fitting Using Random Forest Regression Voting

Cootes, Timothy F.; Ionita, Mircea C.; Lindner, Claudia; Sauer, Patrick

doi:10.1007/978-3-642-33786-4_21

Cited by 235 publications

(214 citation statements)

References 29 publications

Supporting

Mentioning

207

Contrasting

Order By: Relevance

“…As pointed by [5], it is important to note that the curves of the works of [27] and [2] are hardly comparable as they re-annotated some of the data. Fig.…”

Section: Results and Comparisonmentioning

confidence: 99%

See 1 more Smart Citation

Multi-Kernel Appearance Model

Rapp

Bailly

Senechal

et al. 2013

Image and Vision Computing

View full text Add to dashboard Cite

“…As pointed by [5], it is important to note that the curves of the works of [27] and [2] are hardly comparable as they re-annotated some of the data. Fig.…”

Section: Results and Comparisonmentioning

confidence: 99%

“…One regressor is tasked with finding the angle of v, and another is to predict the vector length. For [5], the displacement vector is learned using Random Forest regressor. [10] proposes using a GentleBoost algorithm to predict the horizontal and vertical displacement towards the good location.…”

Section: Landmark Detectionmentioning

confidence: 99%

Multi-Kernel Appearance Model

Rapp

Bailly

Senechal

et al. 2013

Image and Vision Computing

View full text Add to dashboard Cite

“…The landmarks were identified as the most identifiable anatomical landmarks in the images. Recent studies have addressed the topic of automatically segment contours from X-Ray images using random forests [22,23]. However, this methods needs to be trained over a consistent number of images to be used with non-standard projections.…”

Section: In Paragraph 22 We Modified the Text As Followsmentioning

confidence: 99%

Fluoroscopy-based tracking of femoral kinematics with statistical shape models

Valenti

Momi

et al. 2015

Int J CARS

View full text Add to dashboard Cite

Abstract:Purpose Precise knee kinematics assessment helps to diagnose knee pathologies and to improve the design of customized prosthetic components. The first step in identifying knee kinematics is to assess the femoral motion in the anatomical frame. However, no work has been done on pathological femurs, whose shape can be highly different from healthy ones. Method We propose a new femoral tracking technique based on Statistical Shape Models (SSMs) and two calibrated fluoroscopic images, taken at different flexion-extension angles. The cost function optimization is based on genetic algorithms, to avoid local minima. The proposed approach was evaluated on 3 sets of digitally reconstructed radiographic images of osteoarthritic patients. Results It is found that using the estimated shape, rather than that calculated from CT, significantly reduces the pose accuracy, but still has reasonably good results (angle errors around 2 degrees, translation around 1.5mm). Cover Letter. Powered by Editorial Manager® and ProduXion Manager® from Aries Systems CorporationThe authors thank the reviewers and the editor for their interest in this topic and for giving them the possibility to improve the paper with their comments. Hereunder we answer the questions arisen, with indications on the changes made to the manuscript According to the reviewer comment, we modified the text in the "State of the Art" section as follows:"Knee kinematics assessment has great importance both to understand the problems associated with a large number of knee pathologies and to improve the design of prosthetic components. In case of severe osteoarthritis, that are eligible for joint implant surgery, in vivo pre-operatory knee kinematics is fundamental to understand the relative motion between the three joint bones. The relative movement can give an insight of how the ligaments are stretched and their stability, and how could be the feeling for the patient. In this way, pre operatory knee kinematics could help the surgeon to decide which prosthesis should be used and how to correct the misalignment of the bones [3]. Currently, the reconstruction of the pose of the knee can be done using 3D scans such as real-time Magnetic Resonance Imaging (MRI) or through a 2D/3D registration method that superimposes the shape extracted from MRI or Computed Tomography (CT) onto an image, usually X-ray or fluoroscopy. Real-time MRI is suitable to study joint kinematics, as it evidences the muscle structure during movements. However, it can only be used with relatively slow movements, and the accuracy obtained increases from 1mm to more than 3mm depending on the velocity of the movement. In addition, MRI scans are highly expensive [4]. Traditional CT and MRI provide an accurate evaluation of the morphology of the knee, but are limited to static positioning of the patient.[…]The use of dynamic fluoroscopy to detect knee kinematics is described in [12]. The authors use a fluoroscopic system flashing at 30Hz, obtaining continuous images of the knee flexion from 0° to 120°...

show abstract

“…Both MRF and tree-structured models encode the shape in pair-wise geometric relations between parts. To leverage other relations, regression-based methods [4,6,11,28,35] directly predicts the shape parameters from the image.…”

Section: Related Workmentioning

confidence: 99%

Exemplar-Based Graph Matching for Robust Facial Landmark Localization

Zhou

Brandt

Lin

2013

2013 IEEE International Conference on Computer Vision

View full text Add to dashboard Cite

Localizing facial landmarks is a fundamental step in facial image analysis. However, the problem is still challenging due to the large variability in pose and appearance, and the existence of occlusions in real-world face images. In this paper, we present exemplar-based graph matching (EGM), a robust framework for facial landmark localization. Compared to conventional algorithms, EGM has three advantages: (1) an affine-invariant shape constraint is learned online from similar exemplars to better adapt to the test face; (2) the optimal landmark configuration can be directly obtained by solving a graph matching problem with the learned shape constraint; (3) the graph matching problem can be optimized efficiently by linear programming. To our best knowledge, this is the first attempt to apply a graph matching technique for facial landmark localization. Experiments on several challenging datasets demonstrate the advantages of EGM over state-of-the-art methods.

show abstract

Robust and Accurate Shape Model Fitting Using Random Forest Regression Voting

Cited by 235 publications

References 29 publications

Multi-Kernel Appearance Model

Multi-Kernel Appearance Model

Fluoroscopy-based tracking of femoral kinematics with statistical shape models

Exemplar-Based Graph Matching for Robust Facial Landmark Localization

Contact Info

Product

Resources

About