-
Owen Carmichael and
Martial Hebert.
Shape-based Recognition Of Wiry Objects.
In IEEE Conference On Computer Vision And Pattern Recognition,
June 2003.
IEEE Press.
(url)
(pdf)
Keywords:
object recognition,
computer vision.
Abstract: "We present an approach to the recognition of complex-shaped objects in cluttered environments based on edge cues. We first use example images of the desired object in typical backgrounds to train a classifier cascade which determines whether edge pixels in an image belong to an instance of the object or the clutter. Presented with a novel image, we use the cascade to discard clutter edge pixels. The features used for this classification are localized, sparse edge density operations. Experiments validate the effectiveness of the technique for recognition of complex objects in cluttered indoor scenes under arbitrary out-of-image-plane rotation."
@inproceedings{Carmichael_2003_4386,
author = "Owen Carmichael and Martial Hebert",
title = "Shape-based Recognition Of Wiry Objects",
booktitle = "IEEE Conference On Computer Vision And Pattern Recognition",
month = "June",
year = "2003",
publisher = "IEEE Press",
pdf = "http://www.ri.cmu.edu/pub_files/pub4/carmichael_owen_2003_1/carmichael_owen_2003_1.pdf",
url="http://www.ri.cmu.edu/pubs/pub_4386.html",
abstract="We present an approach to the recognition of complex-shaped objects in cluttered environments based on edge cues. We first use example images of the desired object in typical backgrounds to train a classifier cascade which determines whether edge pixels in an image belong to an instance of the object or the clutter. Presented with a novel image, we use the cascade to discard clutter edge pixels. The features used for this classification are localized, sparse edge density operations. Experiments validate the effectiveness of the technique for recognition of complex objects in cluttered indoor scenes under arbitrary out-of-image-plane rotation.",
keywords="object recognition, computer vision"
}
-
Cristian Dima,
Nicolas Vandapel, and
Martial Hebert.
Sensor and Classifier Fusion for Outdoor Obstacle Detection: an Application of Data Fusion To Autonomous Off-Road Navigation.
In The 32nd Applied Imagery Recognition Workshop (AIPR2003),
October 2003.
IEEE Computer Society.
(pdf)
@inproceedings{Dima_2003_4582,
author = "Cristian Dima and Nicolas Vandapel and Martial Hebert",
title = "Sensor and Classifier Fusion for Outdoor Obstacle Detection: an Application of Data Fusion To Autonomous Off-Road Navigation",
booktitle = "The 32nd Applied Imagery Recognition Workshop (AIPR2003)",
month = "October",
year = "2003",
publisher = "IEEE Computer Society",
pdf ="http://www.ri.cmu.edu/pub_files/pub4/dima_cristian_2003_1/dima_cristian_2003_1.pdf"
}
-
C. Gordon,
F. Boukamp,
Daniel Huber,
Edward Latimer,
K. Park, and
B. Akinci.
Combining Reality Capture Technologies for Construction Defect Detection: A Case Study.
In EIA9: E-Activities and Intelligent Support in Design and the Built Environment, 9th EuropIA International Conference,
pages 99-108,
October 2003.
Keywords:
3-D perception,
geometric modeling,
object recognition,
defect detection,
construction site modeling.
@inproceedings{Gordon_2003_4670,
author = "C. Gordon and F. Boukamp and Daniel Huber and Edward Latimer and K. Park and B. Akinci",
title = "Combining Reality Capture Technologies for Construction Defect Detection: A Case Study",
booktitle = "EIA9: E-Activities and Intelligent Support in Design and the Built Environment, 9th EuropIA International Conference",
month = "October",
year = "2003",
pages = "99-108",
keywords="3-D perception, geometric modeling, object recognition, defect detection, construction site modeling"
}
-
Daniel Huber and
Martial Hebert.
3-D Modeling Using a Statistical Sensor Model and Stochastic Search.
In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR),
pages 858-865,
June 2003.
(pdf)
Keywords:
3-D perception,
geometric modeling,
3-D modeling,
registration,
surface matching,
automatic modeling.
@inproceedings{Huber_2003_4427,
author = "Daniel Huber and Martial Hebert",
title = "3-D Modeling Using a Statistical Sensor Model and Stochastic Search",
booktitle = "Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)",
month = "June",
year = "2003",
pages = "858-865",
pdf ="http://www.ri.cmu.edu/pub_files/pub4/huber_daniel_2003_2/huber_daniel_2003_2.pdf",
keywords="3-D perception, geometric modeling, 3-D modeling, registration, surface matching, automatic modeling"
}
-
Daniel Huber and
Nicolas Vandapel.
Automatic 3-D underground mine mapping.
In International Conference on Field and Service Robotics,
July 2003.
Keywords:
3-D perception,
geometric modeling,
3-D modeling,
registration,
mine mapping.
@inproceedings{Huber_2003_4414,
author = "Daniel Huber and Nicolas Vandapel",
title = "Automatic 3-D underground mine mapping",
booktitle = "International Conference on Field and Service Robotics",
month = "July",
year = "2003",
keywords="3-D perception, geometric modeling, 3-D modeling, registration, mine mapping"
}
-
Alonzo Kelly and
Ranjith Unnikrishnan.
Efficient Construction of Globally Consistent Ladar Maps using Pose Network Topology and Nonlinear Programming.
In Proceedings of the 11th International Symposium of Robotics Research (ISRR '03),
November 2003.
@inproceedings{Kelly_2003_4587,
author = "Alonzo Kelly and Ranjith Unnikrishnan",
title = "Efficient Construction of Globally Consistent Ladar Maps using Pose Network Topology and Nonlinear Programming",
booktitle = "Proceedings of the 11th International Symposium of Robotics Research (ISRR '03)",
month = "November",
year = "2003"
}
-
Sanjiv Kumar and
Martial Hebert.
Discriminative Fields for Modeling Spatial Dependencies in Natural Images.
In in proc. advances in Neural Information Processing Systems (NIPS),
December 2003.
(pdf)
@inproceedings{Kumar_2003_4597,
author = "Sanjiv Kumar and Martial Hebert",
title = "Discriminative Fields for Modeling Spatial Dependencies in Natural Images",
booktitle = "in proc. advances in Neural Information Processing Systems (NIPS)",
month = "December",
year = "2003",
pdf = "http://www.ri.cmu.edu/pub_files/pub4/kumar_sanjiv_2003_3/kumar_sanjiv_2003_3.pdf"
}
-
Sanjiv Kumar and
Martial Hebert.
Man-Made Structure Detection in Natural Images using a Causal Multiscale Random Field.
In in proc. IEEE International Conference on Computer Vision and Pattern Recognition (CVPR),
volume 1,
pages 119-126,
2003.
(pdf)
@inproceedings{Kumar_2003_4595,
author = "Sanjiv Kumar and Martial Hebert",
title = "Man-Made Structure Detection in Natural Images using a Causal Multiscale Random Field",
booktitle = "in proc. IEEE International Conference on Computer Vision and Pattern Recognition (CVPR)",
year = "2003",
volume = "1",
pages = "119-126",
pdf ="http://www.ri.cmu.edu/pub_files/pub4/kumar_sanjiv_2003_2/kumar_sanjiv_2003_2.pdf"
}
-
Sanjiv Kumar and
Martial Hebert.
Discriminative Random Fields: A Discriminative Framework for Contextual Interaction in Classification.
In Proceedings of the 2003 IEEE International Conference on Computer Vision (ICCV '03),
volume 2,
pages 1150-1157,
2003.
(pdf)
@inproceedings{Kumar_2003_4596,
author = "Sanjiv Kumar and Martial Hebert",
title = "Discriminative Random Fields: A Discriminative Framework for Contextual Interaction in Classification",
booktitle = "Proceedings of the 2003 IEEE International Conference on Computer Vision (ICCV '03)",
year = "2003",
volume = "2",
pages = "1150-1157",
pdf ="http://www.ri.cmu.edu/pub_files/pub4/kumar_sanjiv_2003_4/kumar_sanjiv_2003_4.pdf"
}
-
Shyjan Mahamud and
Martial Hebert.
The Optimal Distance Measure for Object Detection.
In IEEE Conf. on Computer Vision and Pattern Recognition (CVPR),
2003.
(url)
(pdf)
Abstract: "We develop a multi-class object detection framework whose core component is a nearest neighbor search over object part classes. The performance of the overall system is critically dependent on the distance measure used in the nearest neighbor search. A distance measure that minimizes the mis-classification risk for the 1-nearest neighbor search can be shown to be the probability that a pair of input image measurements belong to different classes. In practice, we model the optimal distance measure using a linear logistic model that combines the discriminative powers of more elementary distance measures associated with a collection of simple to construct feature spaces like color, texture and local shape properties. Furthermore, in order to perform search over large training sets efficiently, the same framework was extended to find hamming distance measures associated with simple discriminators. By combining this discrete distance model with the continuous model, we obtain a hierarchical distance model that is both fast and accurate. Finally, the nearest neighbor search over object part classes was integrated into a whole object detection system and evaluated against an indoor detection task yielding good results."
@inproceedings{Mahamud_2003_4708,
author = "Shyjan Mahamud and Martial Hebert",
title = "The Optimal Distance Measure for Object Detection",
booktitle = "IEEE Conf. on Computer Vision and Pattern Recognition (CVPR)",
year = "2003",
pdf="http://www.ri.cmu.edu/pub_files/pub4/mahamud_shyjan_2003_3/mahamud_shyjan_2003_3.pdf",
url="http://www.ri.cmu.edu/pubs/pub_4708.html",
abstract="We develop a multi-class object detection framework whose core component is a nearest neighbor search over object part classes. The performance of the overall system is critically dependent on the distance measure used in the nearest neighbor search. A distance measure that minimizes the mis-classification risk for the 1-nearest neighbor search can be shown to be the probability that a pair of input image measurements belong to different classes. In practice, we model the optimal distance measure using a linear logistic model that combines the discriminative powers of more elementary distance measures associated with a collection of simple to construct feature spaces like color, texture and local shape properties. Furthermore, in order to perform search over large training sets efficiently, the same framework was extended to find hamming distance measures associated with simple discriminators. By combining this discrete distance model with the continuous model, we obtain a hierarchical distance model that is both fast and accurate. Finally, the nearest neighbor search over object part classes was integrated into a whole object detection system and evaluated against an indoor detection task yielding good results."
}
-
Shyjan Mahamud and
Martial Hebert.
Minimum Risk Distance Measure for Object Recognition.
In IEEE International Conference on Computer Vision (ICCV),
2003.
(url)
(pdf)
Abstract: "Recently, the optimal distance measure for a given object discrimination task under the nearest neighbor framework was derived. For ease of implementation and efficiency considerations, the optimal distance measure was approximated by combining more elementary distance measures defined on simple feature spaces. In this paper, we address two important issues that arise in practice for such an approach: (a) What form should the elementary distance measure in each feature space take? We motivate the need to use optimal distance measures in simple feature spaces as the elementary distance measures; such distance measures have the desirable property that they are invariant to distance-respecting transformations. (b) How do we combine the elementary distance measures? We present the precise statistical assumptions under which a linear logistic model holds exactly. We benchmark our model with three other methods on a challenging face discrimination task and show that our approach is competitive with the state of the art."
@inproceedings{Mahamud_2003_4706,
author = "Shyjan Mahamud and Martial Hebert",
title = "Minimum Risk Distance Measure for Object Recognition",
booktitle = "IEEE International Conference on Computer Vision (ICCV)",
year = "2003",
pdf="http://www.ri.cmu.edu/pub_files/pub4/mahamud_shyjan_2003_1/mahamud_shyjan_2003_1.pdf",
url="http://www.ri.cmu.edu/pubs/pub_4706.html",
abstract="Recently, the optimal distance measure for a given object discrimination task under the nearest neighbor framework was derived. For ease of implementation and efficiency considerations, the optimal distance measure was approximated by combining more elementary distance measures defined on simple feature spaces. In this paper, we address two important issues that arise in practice for such an approach: (a) What form should the elementary distance measure in each feature space take? We motivate the need to use optimal distance measures in simple feature spaces as the elementary distance measures; such distance measures have the desirable property that they are invariant to distance-respecting transformations. (b) How do we combine the elementary distance measures? We present the precise statistical assumptions under which a linear logistic model holds exactly. We benchmark our model with three other methods on a challenging face discrimination task and show that our approach is competitive with the state of the art."
}
-
Aaron Christopher Morris,
Raghavendra Rao Donamukkala,
Anuj Kapuria,
Aaron M Steinfeld,
J. Matthews,
J. Dunbar-Jacobs, and
Sebastian Thrun.
Robotic Walker that Provides Guidance.
In Proceedings of the 2003 IEEE Conference on Robotics and Automation (ICRA '03),
May 2003.
@inproceedings{Morris_2003_4476,
author = "Aaron Christopher Morris and Raghavendra Rao Donamukkala and Anuj Kapuria and Aaron M Steinfeld and J. Matthews and J. Dunbar-Jacobs and Sebastian Thrun",
title = "Robotic Walker that Provides Guidance",
booktitle = "Proceedings of the 2003 IEEE Conference on Robotics and Automation (ICRA '03)",
month = "May",
year = "2003"
}
-
Aaron Christopher Morris,
Derek Kurth,
Daniel Huber,
Chuck Whittaker, and
Scott Thayer.
Case Studies of a Borehole Deployable Robot for Limestone Mine Profiling and Mapping.
In International Conference on Field and Service Robotics (FSR),
July 2003.
Keywords:
3-D perception,
geometric modeling,
3-D modeling,
registration,
mine mapping.
@inproceedings{Morris_2003_4428,
author = "Aaron Christopher Morris and Derek Kurth and Daniel Huber and Chuck Whittaker and Scott Thayer",
title = "Case Studies of a Borehole Deployable Robot for Limestone Mine Profiling and Mapping",
booktitle = "International Conference on Field and Service Robotics (FSR)",
month = "July",
year = "2003",
keywords="3-D perception, geometric modeling, 3-D modeling, registration, mine mapping"
}
-
Bart Nabbe and
Martial Hebert.
Where and When to Look.
In IROS 2003,
October 2003.
IEEE.
(pdf)
Keywords:
outdoor navigation,
dynamic planning,
mid-range sensing,
wide baseline stereo.
@inproceedings{Nabbe_2003_4581,
author = "Bart Nabbe and Martial Hebert",
title = "Where and When to Look",
booktitle = "IROS 2003",
month = "October",
year = "2003",
publisher = "IEEE",
pdf="http://www.ri.cmu.edu/pub_files/pub4/nabbe_bart_2003_1/nabbe_bart_2003_1.pdf",
keywords="outdoor navigation, dynamic planning, mid-range sensing, wide baseline stereo"
}
-
Caroline Pantofaru,
Ranjith Unnikrishnan, and
Martial Hebert.
Toward Generating Labeled Maps from Color and Range Data for Robot Navigation.
In Proceedings of the 2003 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS),
October 2003.
(pdf)
Keywords:
3-D data,
object recognition,
image segmentation,
sensor fusion,
scene understanding.
Abstract: "This paper addresses the problem of extracting information from range and color data acquired by a mobile robot in urban environments. Our approach extracts geometric structures from clouds of 3-D points and regions from the corresponding color images, labels them based on prior models of the objects expected in the environment - buildings in the current experiments - and combines the two sources of information into a composite labeled map. Ultimately, our goal is to generate maps that are segmented into objects of interest, each of which is labeled by its type, e.g., buildings, vegetation, etc. Such a map provides a higher-level representation of the environment than the geometric maps normally used for mobile robot navigation. The techniques presented here are a step toward the automatic construction of such labeled maps."
@inproceedings{pantofaru-iros-03,
author = "Caroline Pantofaru and Ranjith Unnikrishnan and Martial Hebert",
title = "Toward Generating Labeled Maps from Color and Range Data for Robot Navigation",
booktitle = "Proceedings of the 2003 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)",
month = "October",
year = "2003",
pdf = "http://www.ri.cmu.edu/pub_files/pub4/pantofaru_caroline_2003_1/pantofaru_caroline_2003_1.pdf",
keywords="3-D data, object recognition, image segmentation, sensor fusion, scene understanding",
abstract="This paper addresses the problem of extracting information from range and color data acquired by a mobile robot in urban environments. Our approach extracts geometric structures from clouds of 3-D points and regions from the corresponding color images, labels them based on prior models of the objects expected in the environment - buildings in the current experiments - and combines the two sources of information into a composite labeled map. Ultimately, our goal is to generate maps that are segmented into objects of interest, each of which is labeled by its type, e.g., buildings, vegetation, etc. Such a map provides a higher-level representation of the environment than the geometric maps normally used for mobile robot navigation. The techniques presented here are a step toward the automatic construction of such labeled maps."
}
-
Henry Schneiderman.
Learning Statistical Structure for Object Detection.
In Computer Analysis of Images and Patterns (CAIP), 2003,
August 2003.
Springer-Verlag.
(url)
Abstract: "Many classes of images exhibit sparse structuring of statistical dependency. Each variable has strong statistical dependency with a small number of other variables and negligible dependency with the remaining ones. Such structuring makes it possible to construct a powerful classifier by only representing the stronger dependencies among the variables. In particular, a semi-naïve Bayes classifier compactly represents sparseness. A semi-naïve Bayes classifier decomposes the input variables into subsets and represents statistical dependency within each subset, while treating the subsets as statistically inde-pendent. However, learning the structure of a semi-naïve Bayes classifier is known to be NP complete. The high dimensionality of images makes statistical structure learning especially challenging. This paper describes an algorithm that searches for the structure of a semi-naïve Bayes classifier in this large space of possible structures. The algorithm seeks to optimize two cost functions: a localized error in the log-likelihood ratio function to restrict the structure and a global classification error to choose the final structure. We use this approach to train detectors for several objects including faces, eyes, ears, telephones, push-carts, and door-handles. These detectors perform robustly with a high detection rate and low false alarm rate in unconstrained settings over a wide range of variation in background scenery and lighting."
@inproceedings{Schneiderman_2003_4413,
author = "Henry Schneiderman",
title = "Learning Statistical Structure for Object Detection",
booktitle = "Computer Analysis of Images and Patterns (CAIP), 2003",
month = "August",
year = "2003",
publisher = "Springer-Verlag",
url ="http://www.ri.cmu.edu/pubs/pub_4413.html",
abstract="Many classes of images exhibit sparse structuring of statistical dependency. Each variable has strong statistical dependency with a small number of other variables and negligible dependency with the remaining ones. Such structuring makes it possible to construct a powerful classifier by only representing the stronger dependencies among the variables. In particular, a semi-naïve Bayes classifier compactly represents sparseness. A semi-naïve Bayes classifier decomposes the input variables into subsets and represents statistical dependency within each subset, while treating the subsets as statistically inde-pendent. However, learning the structure of a semi-naïve Bayes classifier is known to be NP complete. The high dimensionality of images makes statistical structure learning especially challenging. This paper describes an algorithm that searches for the structure of a semi-naïve Bayes classifier in this large space of possible structures. The algorithm seeks to optimize two cost functions: a localized error in the log-likelihood ratio function to restrict the structure and a global classification error to choose the final structure. We use this approach to train detectors for several objects including faces, eyes, ears, telephones, push-carts, and door-handles. These detectors perform robustly with a high detection rate and low false alarm rate in unconstrained settings over a wide range of variation in background scenery and lighting.",
url="http://www.ri.cmu.edu/pubs/pub_4413.html"
}
-
Ranjith Unnikrishnan and
Martial Hebert.
Robust Extraction of Multiple Structures from Non-uniformly Sampled Data.
In Proceedings of the 2003 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS '03),
volume 2,
pages 1322-29,
October 2003.
(pdf)
Keywords:
mobile robot,
3-D data,
object recognition,
nonparametric statistics,
robust estimation,
scene understanding.
Abstract: "The extraction of multiple coherent structures from point clouds is crucial to the problem of scene modeling. While many statistical methods exist for robust estimation from noisy data, they are inadequate for addressing issues of scale, semi-structured clutter, and large point density variation together with the computational restrictions of autonomous navigation. This paper extends an approach of nonparametric projection-pursuit based regression to compensate for the non-uniform and directional nature of data sampled in outdoor environments. The proposed algorithm is employed for extraction of planar structures and clutter grouping. Results are shown for scene abstraction of 3-D range data in large urban scenes."
@inproceedings{Unnikrishnan_2003_4589,
author = "Ranjith Unnikrishnan and Martial Hebert",
title = "Robust Extraction of Multiple Structures from Non-uniformly Sampled Data",
booktitle = "Proceedings of the 2003 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS '03)",
month = "October",
year = "2003",
volume = "2",
pages = "1322-29",
keywords = "mobile robot",
pdf ="http://www.ri.cmu.edu/pub_files/pub4/unnikrishnan_ranjith_2003_1/unnikrishnan_ranjith_2003_1.pdf",
keywords = {3-D data, object recognition, nonparametric statistics, robust estimation, scene understanding},
abstract="The extraction of multiple coherent structures from point clouds is crucial to the problem of scene modeling. While many statistical methods exist for robust estimation from noisy data, they are inadequate for addressing issues of scale, semi-structured clutter, and large point density variation together with the computational restrictions of autonomous navigation. This paper extends an approach of nonparametric projection-pursuit based regression to compensate for the non-uniform and directional nature of data sampled in outdoor environments. The proposed algorithm is employed for extraction of planar structures and clutter grouping. Results are shown for scene abstraction of 3-D range data in large urban scenes."
}
-
Nicolas Vandapel,
Raghavendra Rao Donamukkala, and
Martial Hebert.
Experimental Results in Using Aerial LADAR Data for Mobile Robot Navigation.
In International Conference on Field and Service Robotics,
2003.
(pdf)
@inproceedings{vandapel-fsr-03,
author = "Nicolas Vandapel and Raghavendra Rao Donamukkala and Martial Hebert",
title = "Experimental Results in Using Aerial LADAR Data for Mobile Robot Navigation",
booktitle = "International Conference on Field and Service Robotics",
year = "2003",
pdf ="http://www.ri.cmu.edu/pub_files/pub4/vandapel_nicolas_2003_1/vandapel_nicolas_2003_1.pdf"
}
-
Nicolas Vandapel,
Raghavendra Rao Donamukkala, and
Martial Hebert.
Quality Assessment of Traversability Maps from Aerial LIDAR Data for an Unmanned Ground Vehicle.
In International Conference on Intelligent Robots and Systems (IROS),
October 2003.
(pdf)
@inproceedings{vandapel-iros-03,
author = "Nicolas Vandapel and Raghavendra Rao Donamukkala and Martial Hebert",
title = "Quality Assessment of Traversability Maps from Aerial LIDAR Data for an Unmanned Ground Vehicle",
booktitle = "International Conference on Intelligent Robots and Systems (IROS)",
month = "October",
year = "2003",
pdf ="http://www.ri.cmu.edu/pub_files/pub4/vandapel_nicolas_2003_2/vandapel_nicolas_2003_2.pdf"
}
-
Jerome Vignola,
Jean-Francois Lalonde, and
Robert Bergevin.
Progressive Human Skeleton Fitting.
In Proceedings of the 16th Conference on Vision Interface,
2003.
|
Annotation: {This paper proposes a method to fit a skeleton or stick-model to a blob to determine the pose of a person in an image. The input is a binary image representing the silhouette of a person and the ouput is a stick-model coherent with the pose of the person in this image. A torso model is first defined, and is then scaled and fitted to the blob using the distance transform of the original image. Then, the fitting is performed independently for each of the four limbs (two arms, two legs), using again the distance transform. The fact that each limb is fitted independently speeds-up the fitting process, avoiding the combinatorial complexity problems that are frequent with this type of method.} url="http://vision.gel.ulaval.ca/fr/publications/Id_444/PublDetails.php" keywords= "pose recognition" .
|
@InProceedings{vignola-vi-03,
author = {Jerome Vignola and Jean-Francois Lalonde and Robert Bergevin},
title = {Progressive Human Skeleton Fitting},
booktitle = {Proceedings of the 16th Conference on Vision Interface},
year = 2003,
annote = {This paper proposes a method to fit a skeleton or stick-model to a blob to determine the pose of a person in an image. The input is a binary image representing the silhouette of a person and the ouput is a stick-model coherent with the pose of the person in this image. A torso model is first defined, and is then scaled and fitted to the blob using the distance transform of the original image. Then, the fitting is performed independently for each of the four limbs (two arms, two legs), using again the distance transform. The fact that each limb is fitted independently speeds-up the fitting process, avoiding the combinatorial complexity problems that are frequent with this type of method.} url="http://vision.gel.ulaval.ca/fr/publications/Id_444/PublDetails.php" keywords= "pose recognition"
}