Non-metric affinity propagation for unsupervised image categorization

Dueck, Delbert; Frey, Brendan J.

doi:10.1109/iccv.2007.4408853

Cited by 243 publications

(147 citation statements)

References 14 publications

Supporting

Mentioning

146

Contrasting

Order By: Relevance

“…Message passing [5] algorithms have been extensively studied and have demonstrated its capability in finding close to optimal solution of complex optimization problems. Inspired by the message passing-based Affinity Propagation algorithm [5], we model the optimization problem in Equation 4 using a graph on which two types of messages are exchanged between nodes (OGs), namely availability and responsibility.…”

Section: Meta Object-group Discovery Via Message Passingmentioning

confidence: 99%

“…Various clustering methods including K-means, spectral clustering have demonstrated strong competency in handing non-linearly connected data [14,16], which are common in multi-label analysis of images. For the task of discovering Meta Object-groups for efficient image retrieval, literature finds that clustering by message passing [5] offers performance advantages in finding clusters based-on non-metric similarity measures. However, its effect in finding related object-groups in BoW features has not yet been studied.…”

Section: Related Workmentioning

confidence: 99%

“…Inspired by the message passing-based Affinity Propagation algorithm [5], we model the optimization problem in Equation 4 using a graph on which two types of messages are exchanged between nodes (OGs), namely availability and responsibility. Responsibility is sent from an OG to all candidate exemplars, quantifying how well-suited a candidate exemplar would be for an OG.…”

Section: Meta Object-group Discovery Via Message Passingmentioning

confidence: 99%

See 2 more Smart Citations

Dimension reduction with meta object-groups for efficient image retrieval

et al. 2015

View full text Add to dashboard Cite

Bag-of-Word (BoW) has been a prominent form for representing visual content such as image and video in recent years, as a result of its unique capability of characterizing visual content in a picture-level while still preserving part of the object-level information. However, it is also noticed that the dimensionality of BoW is usually as high as a few hundreds or even thousands, posing a serious challenge for any application that requires efficient processing. In this article we propose a dimension reduction method called Meta object-Group Component (MGC) to tackle this problem. MGC aims at discovering the hidden relations of objects by examining the correlations between dimensions in the BoW features and maximizing the relations of the members in a meta object-group. By exchanging message passing between object-groups, meta object-groups are identified for a dataset. A meta object-group does not only contain visually similar objects, but also includes objects that are likely to co-occur with each other. As the meta object groups are obtained, group-specific dimension reduction is performed to obtain denser representations for efficient retrieval. We evaluate the framework on the NUS-Wide image dataset with approximately 270,000 images represented by BoW features, and demonstrate its advantage over existing method.

show abstract

Section: Meta Object-group Discovery Via Message Passingmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Meta Object-group Discovery Via Message Passingmentioning

confidence: 99%

See 1 more Smart Citation

Dimension reduction with meta object-groups for efficient image retrieval

et al. 2015

View full text Add to dashboard Cite

show abstract

“…The Caltech-101 dataset [30], collected by Fei-Fei et al, is used in our experiments of object categorization. Following the setting in [1], we select the same twenty object categories from the Caltech-101 dataset, and randomly pick 30 images from each category to form a set of 600 images. The large and diverse intraclass variations make clustering over the dataset very challenging.…”

Section: Visual Object Categorizationmentioning

confidence: 99%

“…As a result, previous research efforts on developing clustering algorithms mostly focus on dealing with different scenarios or specific applications. In the field of vision research, performing data clustering is essential in addressing various tasks such as object categorization [1,2] or image segmentation [3,4]. Despite the great applicability, a fundamental difficulty hindering the advance of clustering techniques is that the intrinsic cluster structure is not evidently revealed in the feature representation of complex data.…”

Section: Introductionmentioning

confidence: 99%

Clustering Complex Data with Group-Dependent Feature Selection

Lin

Liu

Fuh

2010

Computer Vision – ECCV 2010

View full text Add to dashboard Cite

Abstract. We describe a clustering approach with the emphasis on detecting coherent structures in a complex dataset, and illustrate its effectiveness with computer vision applications. By complex data, we mean that the attribute variations among the data are too extensive such that clustering based on a single feature representation/descriptor is insufficient to faithfully divide the data into meaningful groups. The proposed method thus assumes the data are represented with various feature representations, and aims to uncover the underlying cluster structure. To that end, we associate each cluster with a boosting classifier derived from multiple kernel learning, and apply the cluster-specific classifier to feature selection across various descriptors to best separate data of the cluster from the rest. Specifically, we integrate the multiple, correlative training tasks of the cluster-specific classifiers into the clustering procedure, and cast them as a joint constrained optimization problem. Through the optimization iterations, the cluster structure is gradually revealed by these classifiers, while their discriminant power to capture similar data would be progressively improved owing to better data labeling.

show abstract

Unsupervised classification of dynamic obstacles in urban environments

Katz

Nieto

Nebot

2010

Journal of Field Robotics

View full text Add to dashboard Cite

This paper presents a solution to the problem of unsupervised classification of dynamic obstacles in urban environments. A track-based model is introduced for the integration of two-dimensional laser and vision information that provides a robust spatiotemporal synthesis of the sensed moving obstacles and forms the basis for suitable algorithms to perform unsupervised classification by clustering. This work presents various contributions in order to achieve accurate and efficient performance, initially using laser tracks for classification and then incorporating visual tracks to the model. A procedure is proposed for accurate unsupervised classification of dynamic obstacles using a laser stamp representation of the tracks. Laser data are then integrated with visual information through a single-instance visual stamp representation, which is finally extended using a multipleinstance framework to robustly deal with challenges associated with perception in real-world scenarios. The proposed algorithms are extensively validated with a simulated environment. Experiments with a research vehicle in an urban environment demonstrate the performance of the approach with real data. The experimental results reach an accuracy of more than 92% for obstacle classification, finding the clusters that correspond to the main obstacle classes in the data. C

show abstract

Non-metric affinity propagation for unsupervised image categorization

Cited by 243 publications

References 14 publications

Dimension reduction with meta object-groups for efficient image retrieval

Dimension reduction with meta object-groups for efficient image retrieval

Clustering Complex Data with Group-Dependent Feature Selection

Unsupervised classification of dynamic obstacles in urban environments

Contact Info

Product

Resources

About