Minimally supervised acquisition of 3D recognition models from cluttered images
Published 24 August 2005
Andrea Selinger, Randal C. Nelson
Citations13
Generate an AI Snapshot to get a quick, structured summary of this paper.
Study Snapshot
ObjectiveStudy objective
MethodsResearch methodology
PopulationPopulation studied
Sample sizeSample sizes
OutcomesStudy outcomes here
ResultsStudy results comes here
LimitationsResearch study limitations comes here
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
This paper presents a method for minimally supervised training of a previously developed recognition system from unlabeled and unsegmented imagery and shows that the system can successfully extend an object representation extracted from one black background image to contain object features extracted from unlabed cluttered images and can use the extended representation to improve recognition performance on a test set.
Abstract
Appearance-based object recognition systems rely on training from imagery, which allows the recognition of objects without requiting a 3D geometric model. It has been little explored whether such systems can be trained from imagery that is unlabeled, and whether they can be trained from imagery that is not trivially segmentable. In this paper we present a method for minimally supervised training of a previously developed recognition system from unlabeled and unsegmented imagery. We show that the system can successfully extend an object representation extracted from one black background image to contain object features extracted from unlabeled cluttered images and can use the extended representation to improve recognition performance on a test set.
Keywords
Computer ScienceEngineering
International Journal of Computer VisionSnakes: Active contour models
16,951 Citations1988Michael Kass, Andrew Witkin +1 more
This work uses snakes for interactive interpretation, in which user-imposed constraint forces guide the snake near features of interest, and uses scale-space continuation to enlarge the capture region surrounding a feature.
International Journal of Computer VisionVisual learning and recognition of 3-d objects from appearance
1,891 Citations1995Hiroshi Murase, Shree K. Nayar
A near real-time recognition system with 20 complex objects in the database has been developed and a compact representation of object appearance is proposed that is parametrized by pose and illumination.
CVGIP Image UnderstandingA Fast algorithm for active contours and curvature estimation
1,095 Citations1992Donna J. Williams, Mubarak Shah
A greedy algorithm is presented which retains the improvements of stability, flexibility, and inclusion of hard constraints introduced by dynamic programming but is more than an order of magnitude faster than that approach, being O(nm).
NatureA network that learns to recognize three-dimensional objects
936 Citations1990Tomaso Poggio, Shimon Edelman
A scheme is developed, based on the theory of approximation of multivariate functions, that learns from a small set of perspective views a function mapping any viewpoint to a standard view, and a network equivalent to this scheme will 'recognize' the object on which it was trained from any viewpoint.
Lecture notes in computer scienceUnsupervised Learning of Models for Recognition
599 Citations2000Markus Weber, Max Welling +1 more
A method to learn object class models from unlabeled and unsegmented cluttered cluttered scenes for the purpose of visual object recognition that achieves very good classification results on human faces and rear views of cars.
IEEE Transactions on Pattern Analysis and Machine IntelligenceObject matching using deformable templates
518 Citations1996Anil K. Jain, Yu Zhong +1 more
A Bayesian scheme, which is based on prior knowledge and the edge information in the input image, is employed to find a match between the deformed template and objects in the image and computational efficiency is achieved via a coarse-to-fine implementation of the matching algorithm.
Neural ComputationSEEMORE: Combining Color, Shape, and Texture Histogramming in a Neurally Inspired Approach to Visual Object Recognition
344 Citations1997Bartlett W. Mel
It is concluded that in the light of the vast hardware resources available in the ventral stream of the primate visual system relative to those exercised here, the appealingly simple feature-space conjecture remains worthy of serious consideration as a neurobiological model.
Lecture notes in computer scienceObject recognition using multidimensional receptive field histograms
238 Citations1996Bernt Schiele, James L. Crowley
The mathematical foundations of the technique are described and the results of experiments which compare robustness and recognition rates for different local neighborhood operators and histogram similarity measurements are presented.
Artificial IntelligenceAn active vision architecture based on iconic representations
206 Citations1995Rajesh P. N. Rao, Dana H. Ballard
A general active vision architecture based on efficiently computable iconic representations, which employs two primary visual routines, one for identifying the visual image near the fovea (object identification), and another for locating a stored prototype on the retina (object location).
Using Dynamic Programming For Minimizing The Energy Of Active Contours In The Presence Of Hard Constraints
205 Citations2005Amir A. Amini, S. Tehrani +1 more
The optimization problem is set up as a discrete multi-stage decision process and is solved by a “time-delayed” discrete dynamic programming algorithm, which leads to a stable behavior for the active contours over iterations, in addition to allowing for hard constraints to be enforced on the behavior of the solution.
Probabilistic object recognition using multidimensional receptive field histograms
189 Citations1996Bernt Schiele, James L. Crowley
The method for histogram matching is extended to compute the probability of the presence of an object in an image and shows that receptive field histograms provide a technique for object recognition which is robust, has low computational cost and a computational complexity which is linear with the number of pixels.
Combining greyvalue invariants with local constraints for object recognition
97 Citations1996C. Schmid, Roger Mohr
The method is based on local characteristics which are invariant to similarity transformations in the image and computed at automatically detected keypoints using the greyvalue signal and works on images such as paintings for which geometry based recognition fails.
Computer Vision and Image UnderstandingA Perceptual Grouping Hierarchy for Appearance-Based 3D Object Recognition
83 Citations1999Andrea Selinger, Randal C. Nelson
An appearance-based system that uses four distinct levels of perceptual grouping to represent 3D objects in a form that allows not only recognition, but reasoning about 3D manipulation of a sort that has been supported in the past only by 3D geometric models is described.
Object recognition using appearance-based parts and relations
76 Citations2002Chien-Yuan Huang, Octavia Camps +1 more
This paper proposes a representation using appearance-based parts and relations to overcome problems of recognition of general three-dimensional objects in cluttered scenes and its design has been a stumbling block in achieving success.
Vision ResearchLarge-scale tests of a keyed, appearance-based 3-D object recognition system
59 Citations1998Randal C. Nelson, Andrea Selinger
An appearance-based 3-D object recognition system that avoids some of the problems of previous appearance- based schemes is described and a protocol that permits performance in the presence of quantifiable amounts of clutter and occlusion to be predicted on the basis of simple score statistics derived from clean test images and pure clutter images is established.
Clustering appearances of 3D objects
33 Citations2002Ronen Basri, Dan Roth +1 more
This work introduces a method for unsupervised clustering of images of 3D objects that partitions the images into sets that form smooth and parallel surfaces in this space and uses sequences of images to obtain more reliable clustering.
Neural NetworksUnsupervised visual learning of three-dimensional objects using a modular network architecture
16 Citations1999Hideyuki Ando, Satoshi Suzuki +1 more
A modular network architecture that learns to cluster multiple views of multiple three-dimensional (3D) objects based on a mixture of non-linear autoencoders that is superior to the conventional clustering algorithms, such as the K-means algorithm and the Gaussian mixture model.
Learning 3D recognition models for general objects from unlabeled imagery: an experiment in intelligent brute force
9 Citations2002Randal C. Nelson, Andrea Selinger
The problem of training a general, 3D abject recognition system from unlabeled imagery is explored and critical issues and stumbling blocks associated with minimizing the supervision necessary to train such a system are identified.
Automatic hierarchical classification of silhouettes of 3D objects
9 Citations2002Yoram Gdalyahu, Daphna Weinshall
This work proposes a curve dissimilarity measure which relies upon a novel curve matching syntactic algorithm, and uses it to represent the database as a complete graph, with nodes representing the images and Dissimilarity values assigning weights to the edges.
Generic object detection using model based segmentation
8 Citations2003Zhiqian Wang, J. Ben-Arie
A novel approach for detection and segmentation of generic shapes in cluttered images is presented and it is found that angular representation in multidimensional search corresponds better to Euclidean distance than conventional projection and yields improved classification of noisy shapes.