Efficient Texture Retrieval Using Multiscale Local Extrema Descriptors and Covariance Embedding

Pham, Minh‐Tan

doi:10.1007/978-3-030-11018-5_45

Cited by 5 publications

(6 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The disadvantage of Riemannian distance lies in the computation cost using formula (1) is quite expensive because it is necessary to calculating the eigenvalues of the two covariance matrices at the image matching step. Recently, a number of researchers employ Log-Euclidean embedding approaches [40]- [43] to transform the covariance matrices into the linear space and use Euclidean distance as the similarity measure of two covariance matrices R 1 , R 2 . Log-Euclidean between R 1 and R 2 is defined as:…”

Section: Related Workmentioning

confidence: 99%

CDF Space Covariance Matrix of Gabor Wavelet With Convolutional Neural Network for Texture Recognition

Guan

Yang³

et al. 2019

IEEE Access

View full text Add to dashboard Cite

A novel method for texture image recognition is proposed in this paper. The aim of the proposed method is to represent texture by combining the Gabor wavelet transform and deep learning which are efficient techniques for image analysis. We developed the cumulative distribution function (cdf) space covariance model of Gabor wavelet (CSCM-GW), which can jointly model multivariate data in cdf space, in the Gabor wavelet domain to represent texture. The images having different sizes will be transformed by CSCM-GW into same size covariance matrices. Because CSCM-GW is based on the covariance matrix which belongs to Riemannian space, it has the high computational cost in the recognition phase. Therefore, we proposed a novel method of texture recognition called CSCM-GWF-CNN which uses CNN to project the fused covariance of CSCM-GW into low-dimensional vector space for reducing the computational cost and improving the recognition performance. The experiments on Brodatz (111) and KTH-TIPS2-b texture databases show that the proposed method is efficient for texture representation and outperforms most of the state-of-the-art recognition methods. INDEX TERMS Texture recognition, covariance matrix, Gabor wavelet, multidimensional statistical model, convolutional neural network.

show abstract

Section: Related Workmentioning

confidence: 99%

CDF Space Covariance Matrix of Gabor Wavelet With Convolutional Neural Network for Texture Recognition

Guan

Yang³

et al. 2019

IEEE Access

View full text Add to dashboard Cite

show abstract

“…For that reason, Liu et al [22] adopted a pre-trained CNN to obtain features for the texture retrieval task. Despite all the advances in deep learning, the lack of bigger databases for texture retrieval has allowed traditional techniques to be still important for texture retrieval tasks [46], [47].…”

Section: Introductionmentioning

confidence: 99%

Discriminative Features for Texture Retrieval Using Wavelet Packets

2019

View full text Add to dashboard Cite

Wavelet Packets (WPs) bases are explored seeking new discriminative features for texture indexing. The task of WP feature design is formulated as a learning decision problem by selecting the filter-bank structure of a basis (within a WPs family) that offers an optimal balance between estimation and approximation errors. To address this problem, a computationally efficient algorithm is adopted that uses the tree-structure of the WPs collection and the Kullback-Leibler divergence as a discrimination criterion. The adaptive nature of the proposed solution is demonstrated in synthetic and real data scenarios. With synthetic data, we demonstrate that the proposed features can identify discriminative bands, which is not possible with standard wavelet decomposition. With data with real textures, we show performance improvements with respect to the conventional Wavelet-based decomposition used under the same conditions and model assumptions.INDEX TERMS Texture indexing, wavelet packets, minimum probability of error, complexity regularization, minimum cost tree pruning.

show abstract

“…CNN-VGG19 (Napoletano, 2017), yielded competitive results as compared to handcrafted features, such as DDBTC (Guo et al, 2015). It is also noticeable that our proposed approach outperformed all the methods by obtaining an ARR of 80.36% on Outex, 90.25% on USPtex and 81.02% on Stex datasets, whereas the best reported results in the literature were yielded by MS-LED method (Pham, 2018) with ARR of 76.15%, 89.74% and 79.87%, respectively.…”

Section: Methodsmentioning

confidence: 73%

“…Para comparar nuestro enfoque con otros métodos del estado del arte, utilizamos dos protocolos de evaluaci ón diferentes. El protocolo de evaluaci ón utilizado por Outex, Stex y USPtex es la tasa de recuperaci ón media (ARR, del inglés average return rate) propuesta en (Pham, 2018). Sin embargo, los resultados del estado del arte relativos al conjunto de datos TextileTube se proporcionan en base a precision@k, como se propone en (García-Olalla et al, 2018).…”

Section: Experimentos Y Resultadosunclassified

“…To compare our approach with other state-of-the-art methods, we used two different evaluation protocols. The evaluation protocol used by the Outex, Stex and USPtex is Average retrieval rate (ARR) which was suggested in (Pham, 2018). However, the state-of-the-art results concerning the TextileTube dataset are provided based on precision@k, as proposed in (García-Olalla et al, 2018).…”

Section: Evaluation Metricsmentioning

confidence: 99%

See 1 more Smart Citation

Image feature representation using deep learning for instance search and scene recognition = Representación de características de la imagen utilizando el aprendizaje profundo para la búsqueda de instancias y el reconocimiento de escenas

Saikia¹

View full text Add to dashboard Cite

This thesis investigates the creation of novel algorithms for representing images to address two important areas in the field of computer vision: content-based image retrieval (CBIR) and scene recognition. CBIR can be classified into two types, instance-level retrieval and category-level retrieval, and in this thesis, we address the former. Motivated by our joint work with INCIBE, we build deep learning-based systems that can help Law Enforcement Agencies to match evidences in crime scene investigations, among a wide range of other applications. In particular, we propose two algorithms for CBIR, one based on the colour description of objects and the other one on the texture description of patches on images, and another additional algorithm for scene prediction and retrieval that relies on the combination of local and global scene content.CBIR for instance-level retrieval aims at retrieving images from an image or video database that contain the same object or scene as the one depicted in a query image. We introduce two algorithms to address this task in order to gain robustness against colour and texture variances, respectively. On the one hand, we propose colour neural descriptors that are composed of convolutional neural networks (CNNs) features obtained by combining different colour spaces and colour channels. In contrast to previous works, which rely on fine-tuning pre-trained networks, we compute the proposed descriptors based on the activations generated from a pre-trained CNN without fine-tuning. Also, we take advantage of an object detector to optimize the proposed instance retrieval architecture to generate features at both local and global scales. In addition, we introduce a stride based query expansion technique to retrieve objects from multi-view datasets. Finally, we experimentally demonstrated that the proposed colour neural descriptors obtain state-of-the-art results on the Paris 6K, Revisiting-Paris 6k, INSTRE-M and COIL-100 datasets, with mean average precision of 81.70%, 82.02%, 78.8% and 97.9%, respectively.On the other hand, we focus on the texture properties of images. In crime scene investigations, some clues may come from texture patches of images that do not contain much information about the object contour, like a t-shirt lying on the floor. To define the characteristics of such images, the texture patterns are the prime cues for teriores en tres conjuntos de datos p úblicos: MIT-67 Indoor, NYU-v2 y Hotels-50k. La precisi ón alcanzada (MIT-67 Indoor = 94, 5%, NYU-v2 = 74, 5% y la precisi ón top-1 10, 1% sin oclusi ón y 7,8% con oclusi ón media en el Hotels-50k) demostr ó la eficacia del método propuesto, que también supera significativamente los enfoques del estado del arte existentes.Esta tesis contribuye al desarrollo de métodos para crear descriptores robustos a los cambios de color, textura y punto de vista y presenta marcos para utilizarlos en sistemas CBIR y de reconocimiento de escenas.

show abstract

Efficient Texture Retrieval Using Multiscale Local Extrema Descriptors and Covariance Embedding

Cited by 5 publications

References 43 publications

CDF Space Covariance Matrix of Gabor Wavelet With Convolutional Neural Network for Texture Recognition

CDF Space Covariance Matrix of Gabor Wavelet With Convolutional Neural Network for Texture Recognition

Discriminative Features for Texture Retrieval Using Wavelet Packets

Image feature representation using deep learning for instance search and scene recognition = Representación de características de la imagen utilizando el aprendizaje profundo para la búsqueda de instancias y el reconocimiento de escenas

Contact Info

Product

Resources

About