Decomposition, discovery and detection of visual categories using topic models

@article{Fritz2008DecompositionDA,
  title={Decomposition, discovery and detection of visual categories using topic models},
  author={Mario Fritz and Bernt Schiele},
  journal={2008 IEEE Conference on Computer Vision and Pattern Recognition},
  year={2008},
  pages={1-8}
}
  • Mario Fritz, B. Schiele
  • Published 23 June 2008
  • Computer Science
  • 2008 IEEE Conference on Computer Vision and Pattern Recognition
We present a novel method for the discovery and detection of visual object categories based on decompositions using topic models. The approach is capable of learning a compact and low dimensional representation for multiple visual categories from multiple view points without labeling of the training instances. The learnt object components range from local structures over line segments to global silhouette-like descriptions. This representation can be used to discover object categories in a… 

Figures and Tables from this paper

Sub-Category Optimization through Cluster Performance Analysis for Multi-View Multi-Pose Object Detection
TLDR
A sub-category optimization approach that automatically divides an object category into an appropriate number of sub-categories based on appearance variations and outperforms the state-of-the-art methods is proposed.
Spatial-DiscLDA for visual recognition
TLDR
This paper exploits recent advancement in supervised topic modeling, more particularly, the DiscLDA model for object recognition to extend it to a part based visual representation to automatically identify and model different object parts.
Sub-Category Optimization for Multi-view Multi-pose Object Detection
TLDR
A sub-category optimization approach that automatically divides an object category into an appropriate number of sub-categories based on appearance variation by dividing the sample space by unsupervised clustering based on discriminative image features.
Evaluating multi-class learning strategies in a hierarchical framework for object detection
TLDR
It is shown that sequential training achieves the best trade-off between inference and training times at a comparable detection performance and could thus be used to learn the classes on a larger scale.
Modeling, representing and learning of visual categories
TLDR
A hybrid approach is developed that makes use of different levels of supervision and consequently achieves better performance by considering unannotated data and extracts structures that combine the mentioned representation paradigms in a single approach.
Semantic Image Segmentation and Web-Supervised Visual Learning
TLDR
A compact model is introduced, where each class is modelled by a single histogram of visual-words, in contrast to common nearest-neighbour approaches which model each class by storing exemplar histograms, and segmentation algorithms based on these histograms are introduced.
Evaluating multi-class learning strategies in a generative hierarchical framework for object detection
TLDR
This paper shows how different types of multi-class learning can be done within one generative hierarchical framework and provides a rigorous experimental analysis of various object class learning strategies as the number of classes grows.
Recognizing visual object categories with subspace methods and a learned hierarchical shape vocabulary
TLDR
An approach that combines reconstructive and reconstructive subspace methods for robust object classification in a way that enables near-to-perfect classification performance also in the case when objects during testing time are partly occluded is proposed.
Optimization Framework for Learning a Hierarchical Shape Vocabulary for Object Class Detection
TLDR
The idea of this paper is to cast the vocabulary learning into an optimization framework that iteratively improves the hierarchy as a whole and attains a competitive detection performance with respect to the current state-of-the-art at both, faster inference as well as shorter training times.
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 34 REFERENCES
Discovering objects and their location in images
TLDR
This work treats object categories as topics, so that an image containing instances of several categories is modeled as a mixture of topics, and develops a model developed in the statistical text literature: probabilistic latent semantic analysis (pLSA).
Integrating representative and discriminant models for object category detection
TLDR
A method for object category detection which integrates a generative model with a discriminative classifier, which outperforms previously reported results and exploits the strengths of both original methods, minimizing their weaknesses.
Modeling scenes with local descriptors and latent aspects
TLDR
Probabilistic latent semantic analysis generates a compact scene representation, discriminative for accurate classification, and significantly more robust when less training data are available, and the ability of PLSA to automatically extract visually meaningful aspects is exploited to propose new algorithms for aspect-based image ranking and context-sensitive image segmentation.
Multiple Object Class Detection with a Generative Model
TLDR
The performance of the proposed multi-object class detection approach is competitive to state of the art approaches dedicated to a single object class recognition problem.
An Exemplar Model for Learning Object Classes
TLDR
An exemplar model that can learn and generate a region of interest around class instances in a training set, given only a set of images containing the visual class, which enables the detection of multiple instances of the object class in test images.
Cluster Boosted Tree Classifier for Multi-View, Multi-Pose Object Detection
  • Bo Wu, R. Nevatia
  • Computer Science
    2007 IEEE 11th International Conference on Computer Vision
  • 2007
TLDR
This paper proposes a boosting based learning method, called Cluster Boosted Tree (CBT), to automatically construct tree structured object detectors, and shows that this approach outperforms the state-of-the-art methods.
Sharing Visual Features for Multiclass and Multiview Object Detection
TLDR
A multitask learning procedure, based on boosted decision stumps, that reduces the computational and sample complexity by finding common features that can be shared across the classes (and/or views) and considerably reduce the computational cost of multiclass object detection.
A Bayesian hierarchical model for learning natural scene categories
  • Li Fei-Fei, P. Perona
  • Computer Science
    2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05)
  • 2005
TLDR
This work proposes a novel approach to learn and recognize natural scene categories by representing the image of a scene by a collection of local regions, denoted as codewords obtained by unsupervised learning.
Learning hierarchical models of scenes, objects, and parts
TLDR
Applied to a database of images of isolated objects, the sharing of parts among objects improves detection accuracy when few training examples are available and this hierarchical probabilistic model is extended to scenes containing multiple objects.
Learning object categories from Google's image search
TLDR
A new model, TSI-pLSA, is developed, which extends pLSA (as applied to visual words) to include spatial information in a translation and scale invariant manner, and can handle the high intra-class variability and large proportion of unrelated images returned by search engines.
...
1
2
3
4
...