Clustering appearances of objects under varying illumination conditions

Ho, Jeffrey; Yang, Ming-Husang; Lim, Jongwoo; Lee, Kuang-Chih; Kriegman, David

doi:10.1109/cvpr.2003.1211332

Cited by 410 publications

(333 citation statements)

References 26 publications

Supporting

Mentioning

333

Contrasting

Order By: Relevance

“…Sometimes there is nothing disturbing, interfering background, only an object can be seen in each image [27] [34], but sometimes it is difficult to separate the foreground and the background. However, in our task we used heterogeneous image sets, so some of the objects are in unknown background and some of them have no background at all.…”

Section: Related Workmentioning

confidence: 99%

“…The largest difference between the earlier publication and our suggestion is the usefulness of the solution, because our system can be used in more general cases. The earlier published solution contains only color-based feature extraction methods: 3x3x3, 5x5x5 and 6x6x6 quantized RGB histogram (27,125 and 216 bins) and a 32-, 128-, and 256-cell quantized HMMD (MPEG-7-compatible) histogram [25] (32, 128 and 256 bins). These feature extraction methods are not able to grasp variety of an object type (with different shape and illumination).…”

Section: Related Workmentioning

confidence: 99%

“…not the parts of them, like in [39][2]) or objects [27][3] (can be seen) in the parts of the images. The first task is easier, because there is not required to distinguish important and unimportant parts in the images.…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

MMKK++ algorithm for clustering heterogeneous images into an unknown number of clusters

Papp

Szücs

2018

ELCVIA

View full text Add to dashboard Cite

In this paper we present an automatic clustering procedure with the main aim to predict the number of clusters of unknown, heterogeneous images. We used the Fisher-vector for mathematical representation of the images and these vectors were considered as input data points for the clustering algorithm. We implemented a novel variant of K-means, the kernel K-means++, furthermore the min-max kernel K-means plusplus (MMKK++) as clustering method. The proposed approach examines some candidate cluster numbers and determines the strength of the clustering to estimate how well the data fit into clusters, as well as the law of large numbers was used in order to choose the optimal cluster size. We conducted experiments on four image sets to demonstrate the efficiency of our solution. The first two image sets are subsets of different popular collections; the third is their union; the fourth is the complete Caltech101 image set. The result showed that our approach was able to give a better estimation for the number of clusters than the competitor methods. Furthermore, we defined two new metrics for evaluation of predicting the appropriate cluster number, which are capable of measuring the goodness in a more sophisticated way, instead of binary evaluation.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

MMKK++ algorithm for clustering heterogeneous images into an unknown number of clusters

Papp

Szücs

2018

ELCVIA

View full text Add to dashboard Cite

show abstract

“…An interesting formulation has been presented in [8], where the problem of subspace clustering is transformed into a problem of fitting and manipulating polynomials. Moreover, in [9,10], the spectral analysis of an appropriately defined similarity matrix over the data is used to uncover the underlying low dimensional structures as well as the partition that favors them. Recently, in [11], the use of spectral analysis is combined with a multiscale analysis of the growth rate of the local neighborhoods' eigenvalues, so that the appropriate clustering as well as the model parameters, number and dimensionality of the subspaces, are simultaneously recovered from the data.…”

Section: Related Workmentioning

confidence: 99%

Tangent-based manifold approximation with locally linear models

Karygianni¹,

Frossard²

2014

Signal Processing

View full text Add to dashboard Cite

a b s t r a c tIn this paper, we consider the problem of manifold approximation with affine subspaces. Our objective is to discover a set of low dimensional affine subspaces that represent manifold data accurately while preserving the manifold's structure. For this purpose, we employ a greedy technique that partitions manifold samples into groups, which are approximated by low dimensional subspaces. We start by considering each manifold sample as a different group and we use the difference of local tangents to determine appropriate group mergings. We repeat this procedure until we reach the desired number of sample groups. The best low dimensional affine subspaces corresponding to the final groups constitute our approximate manifold representation. Our experiments verify the effectiveness of the proposed scheme and show its superior performance compared to state-of-the-art methods for manifold approximation.

show abstract

“…For facial expression classification, the commonly used Nearest Neighbor Classifier (NNC) [3] and Nearest Subspace Classifier (NSC) [4] locally identify a test sample based on the smallest residual which is measured by the similarity between test sample and each training sample (NNC) or each facial expression class (NSC). Thus, they do not take consideration of the knowledge of training samples of other facial classes.…”

Section: Introductionmentioning

confidence: 99%

Facial Expression Recognition via Sparse Representation

Zhi

Ruan

Wang

2012

IEICE Trans. Inf. & Syst.

View full text Add to dashboard Cite

SUMMARYA facial components based facial expression recognition algorithm with sparse representation classifier is proposed. Sparse representation classifier is based on sparse representation and computed by L1-norm minimization problem on facial components. The features of "important" training samples are selected to represent test sample. Furthermore, fuzzy integral is utilized to fuse individual classifiers for facial components. Experiments for frontal views and partially occluded facial images show that this method is efficient and robust to partial occlusion on facial images. key words: sparse representation, fuzzy integral, decision level fusion, facial expression recognition

show abstract

Clustering appearances of objects under varying illumination conditions

Abstract: We introduce two appearance-based methods for clustering a set of images of 3-D objects, acquired under varying illumination

Cited by 410 publications

References 26 publications

MMKK++ algorithm for clustering heterogeneous images into an unknown number of clusters

MMKK++ algorithm for clustering heterogeneous images into an unknown number of clusters

Tangent-based manifold approximation with locally linear models

Facial Expression Recognition via Sparse Representation

Contact Info

Product

Resources

About