Image classification by non-negative sparse coding, low-rank and sparse decomposition

Zhang, Chunjie; Liu, Jing; Tian, Qi; Xu, Changsheng; Lu, Hanqing; Ma, Songde

doi:10.1109/cvpr.2011.5995484

Cited by 157 publications

(70 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This process is time-consuming and might increase the redundancy in each sub-dictionary, thus not guaranteeing consistency of sparse codes for signals from the same class. [31] presents an image classification framework by using non-negative sparse coding, low-rank and sparse matrix decomposition. A linear SVM classifier is used for the final classification.…”

Section: Related Workmentioning

confidence: 99%

“…Low-rank matrix recovery, which determines a low-rank data matrix from corrupted input data, has been successfully applied to applications including salient object detection [24], segmentation and grouping [35,13,6], background subtraction [7], tracking [34], and 3D visual recovery [13,31]. However, there is limited work [5,19] using this technique for multi-class classification.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Learning Structured Low-Rank Representations for Image Classification

Zhang

Jiang

Davis

2013

2013 IEEE Conference on Computer Vision and Pattern Recognition

244

206

View full text Add to dashboard Cite

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Learning Structured Low-Rank Representations for Image Classification

Zhang

Jiang

Davis

2013

2013 IEEE Conference on Computer Vision and Pattern Recognition

244

206

View full text Add to dashboard Cite

show abstract

“…These methods include local binary pattern [5], spectral-texture feature [7], census transform-structure feature [1], traditional sparse coding (T SC) [8], the nonnegative sparse coding (NN SC) [12], laplacian sparse coding (LSC) [13], and K-SVD sparse coding [14]. For all the coding models, we use SVM as the classifier.…”

Section: Resultsmentioning

confidence: 99%

Compact Sparse Coding for Ground-Based Cloud Classification

Liu

Zhang

Cao

2015

IEICE Trans. Inf. & Syst.

View full text Add to dashboard Cite

SUMMARY Although sparse coding has emerged as an extremely powerful tool for texture and image classification, it neglects the relationship of coding coefficients from the same class in the training stage, which may cause a decline in the classification performance. In this paper, we propose a novel coding strategy named compact sparse coding for groundbased cloud classification. We add a constraint on coding coefficients into the objective function of traditional sparse coding. In this way, coding coefficients from the same class can be forced to their mean vector, making them more compact and discriminative. Experiments demonstrate that our method achieves better performance than the state-of-the-art methods.

show abstract

“…The core idea is to reconstruct a feature with codewords via resolving a leastsquares-based optimization problem with constraints on the codewords and reconstruction coefficients [Huang et al 2013;Cong et al 2011Cong et al , 2013Shao and Tan 2014]. While there are many variations in the recent literature, such as Laplacian sparse coding [Gao et al 2010], mixture sparse coding [Yang et al 2010], multilayer group sparse coding [Gao et al 2011], and nonnegative sparse coding [Zhang et al 2011], the unified representation of sparse coding can be generally written as Given feature descriptors X, the conventional way to solve Equation (1) is to solve it iteratively by alternately optimizing over U or V while fixing the other one, such as ScSPM (Spatial Pyramid Matching based on sparse coding) [Yang et al 2009]. Another kind of method generates the codebook V by clustering (e.g., k-means) in advance and does not change it when obtains the reconstruction coefficients, such as BOW.…”

Section: Review Of Sparse Codingmentioning

confidence: 99%

“…The Vector of Locally Aggregated Descriptors (VLAD) [Jegou et al 2010[Jegou et al , 2012 can be regarded as a simplified version of Fisher vector coding that does not store second-order information about the features. In addition, sparse coding with different constraints on the codewords has also been widely applied for image classification [Yang et al 2009;Zhang et al 2011].…”

Section: Introductionmentioning

confidence: 99%

Multi-Class Latent Concept Pooling for Computer-Aided Endoscopy Diagnosis

Wang

Cong

Fan

et al. 2017

ACM Trans. Multimedia Comput. Commun. Appl.

View full text Add to dashboard Cite

Successful computer-aided diagnosis systems typically rely on training datasets containing sufficient and richly annotated images. However, detailed image annotation is often time consuming and subjective, especially for medical images, which becomes the bottleneck for the collection of large datasets and then building computer-aided diagnosis systems. In this article, we design a novel computer-aided endoscopy diagnosis system to deal with the multi-classification problem of electronic endoscopy medical records (EEMRs) containing sets of frames, while labels of EEMRs can be mined from the corresponding text records using an automatic text-matching strategy without human special labeling. With unambiguous EEMR labels and ambiguous frame labels, we propose a simple but effective pooling scheme called Multi-class Latent Concept Pooling, which learns a codebook from EEMRs with different classes step by step and encodes EEMRs based on a soft weighting strategy. In our method, a computer-aided diagnosis system can be extended to new unseen classes with ease and applied to the standard single-instance classification problem even though detailed annotated images are unavailable. In order to validate our system, we collect 1,889 EEMRs with more than 59K frames and successfully mine labels for 348 of them. The experimental results show that our proposed system significantly outperforms the state-of-the-art methods. Moreover, we apply the learned latent concept codebook to detect the abnormalities in endoscopy images and compare it with a supervised learning classifier, and the evaluation shows that our codebook learning method can effectively extract the true prototypes related to different classes from the ambiguous data.

show abstract

Image classification by non-negative sparse coding, low-rank and sparse decomposition

Cited by 157 publications

References 22 publications

Learning Structured Low-Rank Representations for Image Classification

Learning Structured Low-Rank Representations for Image Classification

Compact Sparse Coding for Ground-Based Cloud Classification

Multi-Class Latent Concept Pooling for Computer-Aided Endoscopy Diagnosis

Contact Info

Product

Resources

About