Vision and Mobile Robotics Laboratory | Publications
Home | Members | Projects | Publications | Software | Videos | Job opportunities Internal

BACK TO INDEX

Publications of year 2002
Thesis
  1. Daniel Huber. Automatic Three-dimensional Modeling from Reality. PhD thesis, Robotics Institute, Carnegie Mellon University, Pittsburgh, PA, December 2002.
    Keywords: 3-D perception, geometric modeling, registration, surface matching, automatic modeling.
    @phdthesis{Huber_2002_4190,
    author = "Daniel Huber",
    title = "Automatic Three-dimensional Modeling from Reality",
    school = "Robotics Institute, Carnegie Mellon University",
    month = "December",
    year = "2002",
    address = "Pittsburgh, PA",
    keywords="3-D perception, geometric modeling, registration, surface matching, automatic modeling" 
    }

Journal articles or book chapters
  1. Henry Schneiderman and Takeo Kanade. Object Detection Using the Statistics of Parts. International Journal of Computer Vision, 2002. (url)
    Abstract: "In this paper we describe a trainable object detector and its instantiations for detecting faces and cars at any size, location, and pose. To cope with variation in object orientation, the detector uses multiple classifiers, each spanning a different range of orientation. Each of these classifiers determines whether the object is present at a specified size within a fixed-size image window. To find the object at any location and size, these classifiers scan the image exhaustively. Each classifier is based on the statistics of localized parts. Each part is a transform from a subset of wavelet coefficients to a discrete set of values. Such parts are designed to capture various combinations of locality in space, frequency, and orientation. In building each classifier, we gathered the class-conditional statistics of these part values from representative samples of object and non-object images. We trained each classifier to minimize classification error on the training set by using Adaboost with Confidence-Weighted Predictions (Shapire and Singer, 1999). In detection, each classifier computes the part values within the image window and looks up their associated class-conditional probabilities. The classifier then makes a decision by applying a likelihood ratio test. For efficiency, the classifier evaluates this likelihood ratio in stages. At each stage, the classifier compares the partial likelihood ratio to a threshold and makes a decision about whether to cease evaluation ¾ labeling the input as non-object - or to continue further evaluation. The detector orders these stages of evaluation from a low-resolution to a high-resolution search of the image. Our trainable object detector achieves reliable and efficient detection of human faces and passenger cars with out-of-plane rotation."
    @article{Schneiderman_2002_3977,
    author = "Henry Schneiderman and Takeo Kanade",
    title = "Object Detection Using the Statistics of Parts",
    journal = "International Journal of Computer Vision",
    year = "2002",
    url="http://www.ri.cmu.edu/pubs/pub_3977.html",
    abstract="In this paper we describe a trainable object detector and its instantiations for detecting faces and cars at any size, location, and pose. To cope with variation in object orientation, the detector uses multiple classifiers, each spanning a different range of orientation. Each of these classifiers determines whether the object is present at a specified size within a fixed-size image window. To find the object at any location and size, these classifiers scan the image exhaustively. Each classifier is based on the statistics of localized parts. Each part is a transform from a subset of wavelet coefficients to a discrete set of values. Such parts are designed to capture various combinations of locality in space, frequency, and orientation. In building each classifier, we gathered the class-conditional statistics of these part values from representative samples of object and non-object images. We trained each classifier to minimize classification error on the training set by using Adaboost with Confidence-Weighted Predictions (Shapire and Singer, 1999). In detection, each classifier computes the part values within the image window and looks up their associated class-conditional probabilities. The classifier then makes a decision by applying a likelihood ratio test. For efficiency, the classifier evaluates this likelihood ratio in stages. At each stage, the classifier compares the partial likelihood ratio to a threshold and makes a decision about whether to cease evaluation ¾ labeling the input as non-object - or to continue further evaluation. The detector orders these stages of evaluation from a low-resolution to a high-resolution search of the image. Our trainable object detector achieves reliable and efficient detection of human faces and passenger cars with out-of-plane rotation." 
    }

  2. Anthony (Tony) Stentz, Cristian Dima, Carl Wellington, Herman Herman, and David Stager. A System for Semi-Autonomous Tractor Operations. Autonomous Robots, 13(1):87-103, July 2002.
    @article{Stentz_2002_4583,
    author = "Anthony (Tony) Stentz and Cristian Dima and Carl Wellington and Herman Herman and David Stager",
    title = "A System for Semi-Autonomous Tractor Operations",
    journal = "Autonomous Robots",
    month = "July",
    year = "2002",
    volume = "13",
    number = "1",
    pages = "87-103" 
    }

Conference's articles
  1. Owen Carmichael and Martial Hebert. Object Recognition by a Cascade of Edge Probes. In British Machine Vision Conference 2002, volume 1, pages 103-112, September 2002. British Machine Vision Association. (url) (pdf)
    Keywords: object recognition, computer vision.
    Abstract: "We frame the problem of object recognition from edge cues in terms of determining whether individual edge pixels belong to the target object or to clutter, based on the configuration of edges in their vicinity. A classifier solves this problem by computing sparse, localized edge features at image locations determined at training time. In order to save computation and solve the aperture problem, we apply a cascade of these classifiers to the image, each of which computes edge features over larger image regions than its predecessors. Experiments apply this approach to the recognition of real objects with holes and wiry components in cluttered scenes under arbitrary out-of-image-plane rotation."
    @inproceedings{Carmichael_2002_4061,
    author = "Owen Carmichael and Martial Hebert",
    title = "Object Recognition by a Cascade of Edge Probes",
    booktitle = "British Machine Vision Conference 2002",
    month = "September",
    year = "2002",
    volume = "1",
    pages = "103-112",
    publisher = "British Machine Vision Association",
    pdf ="http://www.ri.cmu.edu/pub_files/pub3/carmichael_owen_2002_2/carmichael_owen_2002_2.pdf",
    url="http://www.ri.cmu.edu/pubs/pub_4061.html",
    abstract="We frame the problem of object recognition from edge cues in terms of determining whether individual edge pixels belong to the target object or to clutter, based on the configuration of edges in their vicinity. A classifier solves this problem by computing sparse, localized edge features at image locations determined at training time. In order to save computation and solve the aperture problem, we apply a cascade of these classifiers to the image, each of which computes edge features over larger image regions than its predecessors. Experiments apply this approach to the recognition of real objects with holes and wiry components in cluttered scenes under arbitrary out-of-image-plane rotation.",
    keywords=" object recognition, computer vision" 
    }

  2. Peng Chang, Mei Han, and Yihong Gong. Highlight detection and classification of baseball game video with Hidden Markov Models. In Proceedings of the International Conference on Image Processing (ICIP '02), 2002.
    @inproceedings{Chang_2002_4047,
    author = "Peng Chang and Mei Han and Yihong Gong",
    title = "Highlight detection and classification of baseball game video with Hidden Markov Models",
    booktitle = "Proceedings of the International Conference on Image Processing (ICIP '02)",
    year = "2002" 
    }

  3. Peng Chang and Martial Hebert. Robust tracking and structure from motion through sampling based uncertainty representation. In Proceedings of ICRA '02, May 2002. (pdf)
    @inproceedings{Chang_2002_446,
    author = "Peng Chang and Martial Hebert",
    title = "Robust tracking and structure from motion through sampling based uncertainty representation",
    booktitle = "Proceedings of ICRA '02",
    month = "May",
    year = "2002",
    pdf ="http://www.ri.cmu.edu/pub_files/pub3/chang_peng_2002_1/chang_peng_2002_1.pdf" 
    }

  4. Cristian Dima and Simon Lacroix. Using Multiple Disparity Hypotheses for Improved Indoor Stereo. In International Conference on Robotics and Automation, May 2002. IEEE.
    Note: This publication is based on work performed at LAAS-CNRS in Toulouse, France (June-August 2001). (pdf)
    @inproceedings{Dima_2002_4682,
    author = "Cristian Dima and Simon Lacroix",
    title = "Using Multiple Disparity Hypotheses for Improved Indoor Stereo",
    booktitle = "International Conference on Robotics and Automation",
    month = "May",
    year = "2002",
    publisher = "IEEE",
    note = "This publication is based on work performed at LAAS-CNRS in Toulouse, France (June-August 2001).",
    pdf ="http://www.ri.cmu.edu/pub_files/pub4/dima_cristian_2002_1/dima_cristian_2002_1.pdf" 
    }

  5. Martial Hebert, Nicolas Vandapel, Stefan Keller, and Raghavendra Rao Donamukkala. Evaluation and Comparison of Terrain Classification Techniques from LADAR Data for Autonomous Navigation. In 23rd Army Science Conference, December 2002.
    @inproceedings{Hebert_2002_4108,
    author = "Martial Hebert and Nicolas Vandapel and Stefan Keller and Raghavendra Rao Donamukkala",
    title = "Evaluation and Comparison of Terrain Classification Techniques from LADAR Data for Autonomous Navigation",
    booktitle = "23rd Army Science Conference",
    month = "December",
    year = "2002",
    
    }

  6. Sanjiv Kumar, Alex C. Loui, and Martial Hebert. Probabilistic Classification of Image Regions using an Observation-Constrained Generative Approach. In ECCV Workshop on Generative Models based Vision (GMBV), pages 91 - 99, 2002. (pdf)
    @inproceedings{Kumar_2002_4593,
    author = "Sanjiv Kumar and Alex C. Loui and Martial Hebert",
    title = "Probabilistic Classification of Image Regions using an Observation-Constrained Generative Approach",
    booktitle = "ECCV Workshop on Generative Models based Vision (GMBV)",
    year = "2002",
    pages = "91 - 99",
    pdf ="http://www.ri.cmu.edu/pub_files/pub4/kumar_sanjiv_2002_1/kumar_sanjiv_2002_1.pdf" 
    }

  7. Shyjan Mahamud, Martial Hebert, and John Lafferty. Combining Simple Discriminators for Object Discrimination. In European Conf. on Computer Vision (ECCV), 2002. (url) (pdf)
    Abstract: "We propose to combine simple discriminators for object discrimination under the maximum entropy framework or equivalently under the maximum likelihood framework for the exponential family. The duality between the maximum entropy framework and maximum likelihood framework allows us to relate two selection criteria for the discriminators that were proposed in the literature. We illustrate our approach by combining nearest prototype discriminators that are simple to implement and widely applicable as they can be constructed in any feature space with a distance function. For efficient run-time performance we adapt the work on ``alternating trees'' for multi-class discrimination tasks. We report results on a multi-class discrimination task in which significant gains in performance are seen by combining discriminators under our framework from a variety of easy to construct feature spaces."
    @inproceedings{Mahamud_2002_4709,
    author = "Shyjan Mahamud and Martial Hebert and John Lafferty",
    title = "Combining Simple Discriminators for Object Discrimination",
    booktitle = "European Conf. on Computer Vision (ECCV)",
    year = "2002",
    pdf="http://www.ri.cmu.edu/pub_files/pub4/mahamud_shyjan_2002_1/mahamud_shyjan_2002_1.pdf",
    url="http://www.ri.cmu.edu/pubs/pub_4709.html",
    abstract="We propose to combine simple discriminators for object discrimination under the maximum entropy framework or equivalently under the maximum likelihood framework for the exponential family. The duality between the maximum entropy framework and maximum likelihood framework allows us to relate two selection criteria for the discriminators that were proposed in the literature. We illustrate our approach by combining nearest prototype discriminators that are simple to implement and widely applicable as they can be constructed in any feature space with a distance function. For efficient run-time performance we adapt the work on ``alternating trees'' for multi-class discrimination tasks. We report results on a multi-class discrimination task in which significant gains in performance are seen by combining discriminators under our framework from a variety of easy to construct feature spaces." 
    }

  8. Bart Nabbe and Martial Hebert. Toward Practical Cooperative Stereo for Robotic Colonies. In 2002 IEEE International Conference on Robotics and Automation, volume 4, pages 3328-3335, May 2002. Omnipress. (pdf)
    Keywords: wide baseline stereo, affine invariants, robust epipolar estimation.
    @inproceedings{Nabbe_2002_4060,
    author = "Bart Nabbe and Martial Hebert",
    title = "Toward Practical Cooperative Stereo for Robotic Colonies",
    booktitle = "2002 IEEE International Conference on Robotics and Automation",
    month = "May",
    year = "2002",
    volume = "4",
    number = "2002",
    pages = "3328-3335",
    publisher = "Omnipress",
    pdf ="http://www.ri.cmu.edu/pub_files/pub4/nabbe_bart_2002_1/nabbe_bart_2002_1.pdf",
    keywords="wide baseline stereo, affine invariants, robust epipolar estimation" 
    }

  9. Charles Rosenberg and Martial Hebert. Training Object Detection Models with Weakly Labeled Data. In British Machine Vision Conference, September 2002. (pdf)
    @inproceedings{Rosenberg_2002_4578,
    author = "Charles Rosenberg and Martial Hebert",
    title = "Training Object Detection Models with Weakly Labeled Data",
    booktitle = "British Machine Vision Conference",
    month = "September",
    year = "2002",
    pdf ="http://www.ri.cmu.edu/pub_files/pub4/rosenberg_charles_2002_1/rosenberg_charles_2002_1.pdf" 
    }

  10. Ranjith Unnikrishnan and Alonzo Kelly. Mosaicing Large Cyclic Environments for Visual Navigation in Autonomous Vehicles. In IEEE International Conference on Robotics and Automation, 2002 (ICRA '02), volume 4, pages 4299-4306, May 2002.
    @inproceedings{Unnikrishnan_2002_4044,
    author = "Ranjith Unnikrishnan and Alonzo Kelly",
    title = "Mosaicing Large Cyclic Environments for Visual Navigation in Autonomous Vehicles",
    booktitle = "IEEE International Conference on Robotics and Automation, 2002 (ICRA '02)",
    month = "May",
    year = "2002",
    volume = "4",
    pages = "4299-4306" 
    }

  11. Ranjith Unnikrishnan and Alonzo Kelly. A Constrained Optimization Approach to Globally Consistent Mapping. In 2002 IEEE/RSJ Intl. Conference on Intelligent Robots and Systems (IROS '02), volume 1, pages 564-569, October 2002.
    @inproceedings{Unnikrishnan_2002_4098,
    author = "Ranjith Unnikrishnan and Alonzo Kelly",
    title = "A Constrained Optimization Approach to Globally Consistent Mapping",
    booktitle = "2002 IEEE/RSJ Intl. Conference on Intelligent Robots and Systems (IROS '02)",
    month = "October",
    year = "2002",
    volume = "1",
    pages = "564-569" 
    }

Internal reports
  1. Owen Carmichael. Discriminant Filters for Object Recognition. Technical report CMU-RI-TR-02-09, Robotics Institute, Carnegie Mellon University, Pittsburgh, PA, March 2002. (url) (pdf)
    Keywords: object recognition, computer vision, machine learning.
    Abstract: "This paper presents a technique for using training data to design image filters for appearance-based object recognition. Rather than scanning the image with a single set of filters and using the results to test for the existence of objects, we use many sets of filters and take linear combinations of their outputs. The combining coefficients are optimized in a training phase to encourage discriminability between the filter responses for distinct parts of the object and clutter. Our experiments on three popular filter types show that by using this approach to combine sets of filters whose design parameters vary over a wide range, we can achieve detection performance competitive with that of any individual filter set. This in turn can ease the task of fine-tuning the settings for both the filters and the mechanisms that analyze their outputs."
    @techreport{Carmichael_2002_3954,
    author = "Owen Carmichael",
    title = "Discriminant Filters for Object Recognition",
    institution = "Robotics Institute, Carnegie Mellon University",
    month = "March",
    year = "2002",
    number = "CMU-RI-TR-02-09",
    address = "Pittsburgh, PA",
    abstract="This paper presents a technique for using training data to design image filters for appearance-based object recognition. Rather than scanning the image with a single set of filters and using the results to test for the existence of objects, we use many sets of filters and take linear combinations of their outputs. The combining coefficients are optimized in a training phase to encourage discriminability between the filter responses for distinct parts of the object and clutter. Our experiments on three popular filter types show that by using this approach to combine sets of filters whose design parameters vary over a wide range, we can achieve detection performance competitive with that of any individual filter set. This in turn can ease the task of fine-tuning the settings for both the filters and the mechanisms that analyze their outputs.",
    url="http://www.ri.cmu.edu/pubs/pub_3954.html",
    pdf="http://www.ri.cmu.edu/pub_files/pub3/carmichael_owen_2002_1/carmichael_owen_2002_1.pdf",
    keywords="object recognition, computer vision, machine learning" 
    }

Miscellaneous
  1. Ranjith Unnikrishnan. Globally Consistent Mosaicking for Autonomous Visual Navigation. Master's thesis, Robotics Institute, Carnegie Mellon University, Pittsburgh, PA, September 2002.
    @mastersthesis{Unnikrishnan_2002_4063,
    author = "Ranjith Unnikrishnan",
    title = "Globally Consistent Mosaicking for Autonomous Visual Navigation",
    school = "Robotics Institute, Carnegie Mellon University",
    month = "September",
    year = "2002",
    address = "Pittsburgh, PA" 
    }

BACK TO INDEX


The VMR Lab is part of the Vision and Autonomous Systems Center within the Robotics Institute in the School of Computer Science, Carnegie Mellon University.
This page was generated by a modified version of bibtex2html written by Gregoire Malandain