Collaborative Representation Classification Research Articles

Multimodal spatial-temporal feature representation and its application in action recognition

目的 在人体行为识别研究中，利用多模态方法将深度数据与骨骼数据相融合，可有效提高动作的识别率。针对深度图像信息数据量大、冗余度高等问题，提出一种通过获取关键时程信息动作帧序列降低冗余的算法，即质心运动路径松弛算法，并根据不同模态数据的特点，提出一种新的时空特征表示方法。方法 质心运动路径松弛算法根据质心在相邻帧之间的运动距离，计算图像差分后获得的活跃部分的相似系数，然后剔除掉相似度高的帧，获得足以表达行为的关键时程信息。根据图像动态部分的变化特性、人体各部分在运动中的协同性和局部显著性特征构建一种新的时空特征表示方法。结果 在MSR-Action3D数据集上对本文方法的效果进行验证。在3个子集中进行交叉验证的平均分类识别率为95.743 2%，分别比Multi-fused，CovP3DJ，D3D-LSTM（densely connected 3DCNN and long short-term memory），Joint Subset Selection方法高2.443 2%，4.763 2%，0.343 2%，0.213 2%。本文方法在使用完整数据集的扩展实验中进行交叉验证的分类识别率为93.040 3%，具有很好的鲁棒性。结论 实验结果表明，本文提出的去冗余算法在降低冗余后提升了识别效果，提取的特征之间具有相关性低的特点，在组合识别中具有良好的互补性，有效提高了分类识别的精确度。;Objective Human body motion-related recognition has been developing in the context of computer vision and pattern recognition like auxiliary human-computer interaction，motion analysis，intelligent monitoring，and virtual reality. To obtain two-dimensional information for its behavioral recognition，conventional motion behavior recognition is mainly used the RGB image sequence captured by RGB camera. To improve the ability to detect short-duration fragments，current feature descriptors for RGB image sequences are employed to characterize human behavior，such as histogram of oriented gradient（HOG），histogram of optical flow （HOF），and a three-dimensional feature pyramid. Some researchers are focused on the feature that image depth is insensitive to ambient light since RGB images are oriented to behavior image sequences of objects in terms of two-dimensional information. The depth information of the image is coordinated with the features of RGB image to describe the related behavior. Human behavior recognition-relevant multi-modal method can be used to fuse depth data and skeleton data，which can improve the recognition rate of action effectively. Recent depth map is widely used in relevant to human behavior recognition. But，the collection of depth information data is required to be optimized because of time complexity of feature extraction and space complexity of feature storage. To resolve the problems，we develop an algorithm to optimize frames of the depth map and resource consumption. At the same time，a new representation of motion features is facilitated as well according to the motion information of the centroid. Method First，the temporal feature vector is used in terms of depth map sequence-extracted time sequence information. The centroid motion path relaxation algorithm is used to realize depth image de-duplication and de redundancy，and the skeleton map-extracted spatial structure feature vector from are spliced to form the spatio-temporal feature input. Next，spatial features are extracted in terms of the original skeleton points coordinates-spliced three-channel spatial feature map. Finally，the fusion probability of spatio-temporal features and spatial features is used for classification and recognition. Our centroid motion path relaxation algorithm is focused on the optimization of redundant information，the time complexity of feature extraction，and the space complexity of feature storage. For the skeleton data，the global feature of motion direction is proposed to fully reflect the integrity and coordination of limb movements. The extracted features are concatenated to obtain the spatio-temporal feature vector，and they can be fused and enhanced through the original coordinates of skeleton points-built three-channel spatial feature map. Its effectiveness is verified on the MSR-Action3D dataset. Result The experimental setting 1 demonstrate that it is 0. 826 0% higher than the depth motion map（DMM）-local binary pattern（LBP）algorithm，1. 015 2% higher than DMM-CRC（collaborative representation classifier），3. 450 1% higher than gradient local auto correlation（DMM-GLAC） algorithm，0. 605 8% higher than EigenJoint algorithm，and 0. 605 8% higher than space-time auto correlation of gradient （STACOG）algorithm is 10. 624 5% higher. After removing redundancy，the result of experimental setting 1 is 0. 126 1% higher as well. The cross-validation on experimental setting 2 show that the average classification and recognition rate in the three subsets is 95. 743 2%，2. 443 2% higher than multi-fused method，4. 763 2% higher than CovP3DJ method，0. 343 2% higher than D3D-LSTM method，and 0. 213 2% higher than joint subset selection method. For the overall data set，it is 2. 030 3% higher than low latency method，0. 240 3% higher than combination of deep models method，and 2. 340 3% higher than complex network coding method. The experimental setting 2 illustrates that the average classification recognition rate of cross-validation in three subsets is 95. 743 2%，and the classification recognition rate of the complete dataset is 93. 040 3%. Conclusion Our algorithm proposed can improve the recognition effect based on redundancy-optimized，and the featuresextracted have lower correlation mutually，which can improve the accuracy of classification recognition effectively.

Collaborative representation with hole filling techniques for Kinect RGB-D face recognition

We present in this paper, three different filtering techniques based on kernel function to fill missing information in depth map images obtained from low resolution sensor such as Kinect to improve the performance accuracy of RGB-D face recognition systems. We propose in this study, an RGB-D face recognition scheme that combines depth map and colour image using wavelet average fusion followed by collaborative representation classifier (CRC) for comparison of reference and probe images. We present the evaluation results based on our GU-RGBD face database and IIIT-D face database to present the significance of our three different filters employed in hole filling. Further, our investigation presents an extensive experimental analysis using eight different feature extraction techniques independently across three different filters to demonstrate the potential of our proposed approach. The proposed approach of hole filling, improves the performance accuracy of RGB-D face recognition as compared to without employing filtering operations.

Probabilistic Collaborative Representation Based Ensemble Learning for Classification of Wetland Hyperspectral Imagery

Protection of wetlands is important for ecosystem in recent years, and the classification of wetland ground cover is the foundation of investigation and protection work. Probabilistic collaborative representation classifier (ProCRC) is one of the best performing classifiers which has been applied in hyperspectral image (HSI) classification. However, its performance is greatly limited for wetland data where spectrums are highly similar. Moreover, the complex distribution of ground objects in wetlands have not been wisely utilized in the classification. In this article the intrinsic mechanism of ProCRC is found and its kernel version is proposed to solve the problems of wetlands classification. Then, a new ensemble learning strategy that considers neighborhood information are proposed, which largely alleviates the problem of sample collection in wetlands. Under the guidance of this strategy, two specific ensemble learning algorithms, i.e., LNE and LNSAE, are proposed. The superiority of proposed methods is validated using three typical HSI data sets of China coastal wetland with few samples.

Collaborative representation of convolutional neural network features to detect artificial ripening of banana using multispectral imaging

With the increasing demand for fruits in the consumer market and to fulfill the needs of consumers, fruits are deliberately resorted to ripen artificially using industrial-grade Calcium Carbide (CaC2) which has carcinogenic properties. The preferences to employ imaging sensing technology across the electromagnetic spectrum have gained noteworthy attention in recent times. This work presents the study to distinguish between natural and artificially ripened bananas using a multispectral imaging approach. Specifically, we carry out the analysis using multispectral images collected for natural and artificially ripened bananas in eight narrow spectrum bands spanning from visible (VIS) to near-infrared (NIR) wavelength. Further, to present the contribution of each individual spectral band, we propose a scheme that extracts the relevant features from “convolution 5 (conv5)” and “fully connected 6 (fc6)” layers of convolutional neural network (CNN) based on AlexNet architecture and processes these features independently with collaborative representation classifier (CRC) in a robust manner. Essentially, we demonstrate the experimental classification accuracy based on multispectral images comprised of 5760 sample images using 10-fold cross-validation. Based on the proposed scheme, the highest average classification accuracy of about 88.82 ± 1.65% is obtained using our proposed approach presenting the significance of our work. Practical applications The practical application of our proposed approach is to distinguish between natural and artificially ripened banana fruits in a market chain. Further, the approach is non-invasive and non-destructive based on multispectral imaging which has the potential to carry out real-time quality analysis of bananas, especially to detect adulteration.

Land Cover Classification from Hyperspectral Images via Local Nearest Neighbor Collaborative Representation with Tikhonov Regularization

The accurate and timely monitoring of land cover types is of great significance for the scientific planning, rational utilization, effective protection and management of land resources. In recent years, land cover classification based on hyperspectral images and the collaborative representation (CR) model has become a hot topic in the field of remote sensing. However, most of the existing CR models do not consider the problem of sample imbalance, which affects the classification performance of CR models. In addition, the Tikhonov regularization term can improve the classification performance of CR models, but greatly increases the computational complexity of CR models. To address the above problems, a local nearest neighbor (LNN) method is proposed in this paper to select the same number of nearest neighbor samples from each nearest class of the test sample to construct a dictionary. This is then introduced into the original collaborative representation classification (CRC) method and CRC with Tikhonov regularization (CRT) for land cover classification, denoted as LNNCRC and LNNCRT, respectively. To verify the effectiveness of the proposed LNNCRC and LNNCRT methods, the classification performance and running time of the proposed methods are compared with those of six popular CR models on a hyperspectral scene with nine land cover types. The experimental results show that the proposed LNNCRT method achieves the best land cover classification performance, and the proposed LNNCRC and LNNCRT methods not only further exclude the interference of irrelevant training samples and classes, but also effectively eliminate the influence of imbalanced training samples, so as to improve the classification performance of CR models and effectively reduce the computational complexity of CR models.

Inductive Latent Space Sparse and Low-rank Subspace Clustering Algorithm

Sparse subspace clustering (SSC) and low-rank representation (LRR) are the most popular algorithms for subspace clustering. However, SSC and LRR are transductive methods and cannot deal with the new data not involved in the training data. When a new data comes, SSC and LRR need to calculate over all the data again, which is a time-consuming thing. On the other hand, for high-dimensional data, dimensionality reduction is firstly performed before running SSC and LRR algorithms which isolate the dimensionality reduction and the following subspace clustering. To overcome these shortcomings, in this paper, two sparse and low-rank subspace clustering algorithms based on simultaneously dimensionality reduction and subspace clustering which can deal with out-of-sample data were proposed. The proposed algorithms divide the whole data set into in-sample data and out-of-sample data. The in-sample data are used to learn the projection matrix and the sparse or low-rank representation matrix in the low-dimensional space. The membership of in-sample data is obtained by spectral clustering. In the low dimensional embedding space, the membership of out of sample data is obtained by collaborative representation classification (CRC). Experimental results on a variety of data sets verify that our proposed algorithms can handle new data in an efficient way.

Weighted Residual Dynamic Ensemble Learning for Hyperspectral Image Classification

Recently, collaborative representation classifiers have been extensively studied as an essential method for hyperspectral image. However, how to comprehensively utilize the classification advantages of multiple collaborative classifiers has not been well investigated. In this paper, two new dynamic ensemble learning methods using local weighted residual (LWR-DEL) and double-weighted residual (DWR-DEL) of multi-collaborative representation classifiers are proposed. First, the dynamic ensemble learning method based on clustering is utilized to introduce prior knowledge for the collaborative representation classifier. Then, with prior knowledge, the local weights of each classifier for a different region of competence are obtained. To consider the global information of hyperspectral data, the K-nearest neighbor (K-NN) algorithm is adopted to achieve validation samples with global information. The global weights for each classifier can be obtained and then used to constrain the locally weighted residuals. Similar to LWR-DEL, the global information is also used to constrain residual, and then double-weighted constrained residual fusion obtains the final classifier result. The effectiveness of the proposed methods is validated using three hyperspectral data sets. The experimental results show that both LWR-DEL and DWR-DEL outperform their single-classifier counterparts. In particular, the proposed methods provide superior performance compared with the state-of-the-art methods.

Human action recognition using a depth sequence key-frames based on discriminative collaborative representation classifier for healthcare analytics

Using deep map sequence to recognize human action is an important research field in computer vision. The traditional deep map-based methods have a lot of redundant information. Therefore, this paper proposes a new deep map sequence feature expression method based on discriminative collaborative representation classifier, which highlights the time sequence of human action features. In this paper, the energy field is established according to the shape and action characteristics of human body to obtain the energy information of human body. Then the energy information is projected onto three orthogonal axes to obtain deep spatialtemporal energy map. Meanwhile, in order to solve the problem of high misclassification probability of similar samples by collaborative representation classifier (CRC), a discriminative CRC (DCRC) is proposed. The classifier takes into account the influence of all training samples and each kind of samples on the collaborative representation coefficient, it obtains the highly discriminative collaborative representation coefficient, and improves the discriminability of similar samples. Experimental results on MSR Action3D data set show that the redundancy of key-frame algorithm is reduced, and the operation efficiency of each algorithm is improved by 20%-30%. The proposed algorithm in this paper reduces the redundant information in deep map sequence and improves the extraction rate of feature map. It not only preserves the spatial information of human action through the energy field, but also records the temporal information of human action in a complete way. What?s more, it still maintains a high recognition accuracy in the action data with temporal information.

Multimodality for Reliable Single Image Based Face Morphing Attack Detection

Face morphing attacks have demonstrated a high vulnerability on human observers and commercial off-the-shelf Face Recognition Systems (FRS), especially in the border control scenario. Therefore, detecting face morphing attacks is paramount to achieving a reliable and secure border control operation. This work presents a novel framework for the Single image-based Morphing Attack Detection (S-MAD) based on the multimodal regions such as eyes, nose, and mouth. Each of these regions is processed using colour scale-space representation on which two different types of features are extracted using Binarised Statistical Image Features (BSIF) and Local Binary Features (LBP) techniques. These features are then fed to the classifiers such as Probabilistic Collaborative Representation Classifier (P-CRC) and Spectral Regression Kernel Discriminant Analysis (SRKDA). Their decisions are combined at score level to make the final decision. Extensive experiments are carried out on three different face morphing datasets to benchmark the performance of the proposed method with the existing methods. Further, the proposed method is benchmarked on the Bologna Online Evaluation Platform (BOEP). Obtained results demonstrate the improved performance of the proposed method over existing state-of-the-art methods.

Diversity-Driven Multikernel Collaborative Representation Ensemble for Hyperspectral Image Classification

Recently, kernel collaborative representation classification (KCRC) has shown its outstanding performance in dealing with the problem of linear inseparability in hyperspectral remote sensing image classification. Meanwhile, ensemble learning has attracted great attention in improving the performance of a single classifier. Aiming at the limitation of a single classifier, bagging algorithm based on KCRC(KCRC-bagging) is presented in this article. The KCRC-bagging method uses bootstrap to increase the diversity of base classifiers, thus improving the classification accuracy and generalization performance. In order to reduce the scale of ensemble, a diversity-driven multikernel collaborative representation classifier ensemble approach (DIV-KCRC) is proposed. DIV-KCRC verifies the effectiveness of the representation classifier with the pair of diversity measures, and classifiers with high accuracy and diversity are selected to improve the classification performance and efficiency of the ensemble system. Three real hyperspectral data sets were applied to prove the validity of the proposed method. The experimental results demonstrate that both KCRC-bagging and DIV-KCRC can yield better classification performance than their corresponding base classifiers. In particular, DIV-KCRC provides more reliable classification results than KCRC-bagging.

Multiview Spatial–Spectral Active Learning for Hyperspectral Image Classification

Supervised classification algorithms on the intricate ground object information of hyperspectral images (HSIs) require a large number of training samples that are annotated manually for model learning. To reduce the labeling cost and improve training sample effectiveness, a multiview spatial–spectral active learning (MVSS-AL) model is proposed in this study. First, a committee model composed of collaborative representation classification is introduced to form a leave-one-class-out (LOCO) multiview strategy, which explores more effective information in the limited training data. Second, the sample query strategy is designed from the perspective of classification confidence (CC) and training contribution (TC). The most inconsistent high-quality samples are screened by making full use of iterative prediction information and spatial–spectral features contained in hyperspectral imagery. Finally, the spatial–spectral LOCO active learning (AL) model obtains target samples through two-layer screening in each iteration and utilizes a support vector machine to obtain the final classification results. The proposed method is tested on three real-world hyperspectral datasets, and the comparison with several novel methods shows that the proposed method is better in the classification performance of restricted sample training.

Exploring 3D Human Action Recognition Using STACOG on Multi-View Depth Motion Maps Sequences.

This paper proposes an action recognition framework for depth map sequences using the 3D Space-Time Auto-Correlation of Gradients (STACOG) algorithm. First, each depth map sequence is split into two sets of sub-sequences of two different frame lengths individually. Second, a number of Depth Motion Maps (DMMs) sequences from every set are generated and are fed into STACOG to find an auto-correlation feature vector. For two distinct sets of sub-sequences, two auto-correlation feature vectors are obtained and applied gradually to -regularized Collaborative Representation Classifier (-CRC) for computing a pair of sets of residual values. Next, the Logarithmic Opinion Pool (LOGP) rule is used to combine the two different outcomes of -CRC and to allocate an action label of the depth map sequence. Finally, our proposed framework is evaluated on three benchmark datasets named MSR-action 3D dataset, DHA dataset, and UTD-MHAD dataset. We compare the experimental results of our proposed framework with state-of-the-art approaches to prove the effectiveness of the proposed framework. The computational efficiency of the framework is also analyzed for all the datasets to check whether it is suitable for real-time operation or not.

Grouped Collaborative Representation for Hyperspectral Image Classification Using a Two-Phase Strategy

This letter proposes a two-phase strategy-based grouped collaborative representation classifier (CRC) for hyperspectral image classification. Specifically, a spectral correlation-based CRC (SCCR) is proposed in the first phase, which considers a regularization term that uses the spectral correlations between the test sample and training samples. The representation coefficient vector generated by SCCR is then transformed into class-specific group weights. In the second phase, we integrate group weights and spectral correlations into the CRC and propose a grouped CRC (GCRC). Experimental results obtained from three real hyperspectral data sets demonstrate that the proposed SCCR and GCRC can provide better classification performance over other state-of-the-art representation-based classifiers.

Cooperative representation of multiscale patch face recognition based on fuzzy decision

The machine learning of small sample size is one of the most challenging problems in face recognition. Multiscale patch cooperative representation for face recognition provides multiple patch scales to a sample set. In each patch scale, all samples in the training set and test set are segmented into patches of the same size, followed by a collaborative representation classification. A sample belongs to the category in which the number of patch blocks is the highest. Under a single scale, however, the patch blocks of a sample often belong to different categories. If the decision value of the sample is defined as the category with the largest number of patches, the possibility of the sample belonging to other categories is often disregarded; this will then significantly reduce the recognition accuracy. Therefore, a multiscale patched collaborative representation based on fuzzy decision is proposed herein. In a single scale, among all the patch blocks in a sample, the proportion of patch blocks belonging to a category to the total patch blocks is used to represent the degree of a sample belonging to the category. Hence, a fuzzy decision matrix can be obtained for a sample set in each scale. The elements of the decision matrix represent the possibility of samples belonging to categories, thereby solving the absolute problem of classification. Corresponding weights are applied to the patch scales, and multiscale outputs can be integrated by regularizing boundary distribution optimization. It is experimentally demonstrated that the proposed method exhibits high recognition accuracy and is superior to some existing algorithms.

Error-Tolerant Deep Learning for Remote Sensing Image Scene Classification.

Due to its various application potentials, the remote sensing image scene classification (RSSC) has attracted a broad range of interests. While the deep convolutional neural network (CNN) has recently achieved tremendous success in RSSC, its superior performances highly depend on a large number of accurately labeled samples which require lots of time and manpower to generate for a large-scale remote sensing image scene dataset. In contrast, it is not only relatively easy to collect coarse and noisy labels but also inevitable to introduce label noise when collecting large-scale annotated data in the remote sensing scenario. Therefore, it is of great practical importance to robustly learn a superior CNN-based classification model from the remote sensing image scene dataset containing non-negligible or even significant error labels. To this end, this article proposes a new RSSC-oriented error-tolerant deep learning (RSSC-ETDL) approach to mitigate the adverse effect of incorrect labels of the remote sensing image scene dataset. In our proposed RSSC-ETDL method, learning multiview CNNs and correcting error labels are alternatively conducted in an iterative manner. It is noted that to make the alternative scheme work effectively, we propose a novel adaptive multifeature collaborative representation classifier (AMF-CRC) that benefits from adaptively combining multiple features of CNNs to correct the labels of uncertain samples. To quantitatively evaluate the performance of error-tolerant methods in the remote sensing domain, we construct remote sensing image scene datasets with: 1) simulated noisy labels by corrupting the open datasets with varying error rates and 2) real noisy labels by deploying the greedy annotation strategies that are practically used to accelerate the process of annotating remote sensing image scene datasets. Extensive experiments on these datasets demonstrate that our proposed RSSC-ETDL approach outperforms the state-of-the-art approaches.

Cross-view kernel collaborative representation classification for person re-identification

Currently, person re-identification (re-ID) has been applied in many public security applications. Yet owing to the big visual appearance changes of the same identity under different views, re-ID still faces many challenges. To reduce the intra-person discrepancy, extracting more power feature representations from pedestrian images is a reasonable solution. We propose a cross-view kernel collaborative representation based classification (CV-KCRC) method for person re-ID in our work. Our method aims to find more robust and discriminative feature representations that embody cross-view information to enhance the identification capability of features. We map the image features into a high dimensional feature space first and then use view-specific projection matrices to project the high dimensional features into a common low dimensional subspace. We expect that in the shared subspace the codings of same person from different views have the highest similarity and better performance can be achieved. Experiments on seven commonly used datasets reveal that our algorithm outperforms many state-of-the-art algorithms.

Hyperspectral face recognition with a spatial information fusion for local dynamic texture patterns and collaborative representation classifier

Hyperspectral face recognition provides improved classification rates due to its abundant information in the face cubes of every subject in hyperspectral face databases. However, while offering excellent opportunities, it also brings new challenges, such as low signal-to-noise ratio, interband misalignment, and high data dimensionality. Based on these ad hoc problems, literature has already proposed some optimisation methods including dimensionality reduction, image denoising, and alignment to perform face recognition, yet lacking comprehensive evaluation. This paper proposes a novel hyperspectral face recognition algorithm that is based on spatial information fusion for feature extraction (histogram of local dynamic texture patterns) and collaborative representation classifier for classification. Meanwhile, the algorithm is applied to three popular hyperspectral face databases, Carnegie Mellon University (CMU)-hyperspectral face database (HSFD), University of Western Australia (UWA)-HSFD, and Hong Kong Polytechnic University (PolyU)-HSFD databases. Experimental results demonstrate that CMU-HSFD and UWA-HSFD databases achieve very competitive classification results. PolyU-HSFD database also achieves rather good classification rates. The best recognition results are 98.5% ± 0.95, 96.6% ± 0.98, and 94.0% ± 2.86 for CMU-HSFD, UWA-HSFD and PolyU-HSFD, respectively. It demonstrates experimentally that this algorithm can be used to recognise faces. Moreover, we compared eight existing state-of-the-art face recognition techniques with our proposed method in performing hyperspectral face recognition. In this research, we formulate hyperspectral face recognition as an image-set classification problem and evaluate the performances compared with other kinds of algorithms. Comparisons with the eight existing hyperspectral face recognition techniques on three standard datasets show that the proposed algorithm outperforms most other state-of-the-art algorithms, indicating that it is a promising approach for hyperspectral face recognition.

A Novel Deep Learning Model for Palmprint/Palmvein Recognition

In this article, a novel palmprint/palmvein recognition algorithm is proposed. The algorithm firstly extracts the palmprint/palmvein feature by the principal component analysis network (PCANet), and then classifies by collaborative representation classifier (CRC). The proposed method is validated on several palmprint and palmvein databases, and the experimental results show that the method is very effective. At the same time, this method has good robustness for small training set. In the datasets of blue illuminations of Hong Kong Polytechnic University (PolyU) multispectral palmprint database, the recognition rate can reach 100% with only one training sample of each class. PCANet can extract palmprint depth feature information without the intervention of prior knowledge, and the CRC classifier can achieve high recognition accuracy rate while having an extremely low computational complexity. This algorithm can be used to real-time application.

Fusing Deep Features by Kernel Collaborative Representation for Remote Sensing Scene Classification

Remote sensing scene classification is widely concerned because of its wide applications. Recently, convolutional neural networks (CNNs) have made a significant breakthrough in remote sensing image scene classification. However, the accuracy of using only fully connected layer of CNNs as a classifier is not satisfied, especially for few-shot remote sensing images. In this paper, we propose a framework of Feature-Fusion based Kernel Collaborative Representation Classification (FF-KCRC) for few-shot remote sensing images, which can make full use of the synergy between samples and the similarity between different types of image features to improve the accuracy of scene classification for few-shot remote sensing images. Specifically, we first design an effective feature extraction strategy to obtain more discriminative image features from CNNs, in which transfer learning is used to transfer the weights of pre-trained CNNs to alleviate the few-shot training problem. In the next, we design the FF-KCRC framework to make full use of the synergy between different categories and fuse the classification of different features, where kernel trick is used to address the problem of the linearly indivisibility. Extensive experiments have been conducted on publicly available remote sensing image data sets and the results show that the proposed FF-KCRC achieves stateof-the-art results.

NSCR-Based DenseNet for Lung Tumor Recognition Using Chest CT Image.

Nonnegative sparse representation has become a popular methodology in medical analysis and diagnosis in recent years. In order to resolve network degradation, higher dimensionality in feature extraction, data redundancy, and other issues faced when medical images parameters are trained using convolutional neural networks. Lung tumors in chest CT image based on nonnegative, sparse, and collaborative representation classification of DenseNet (DenseNet-NSCR) are proposed by this paper: firstly, initialization parameters of pretrained DenseNet model using transfer learning; secondly, training DenseNet using CT images to extract feature vectors for the full connectivity layer; thirdly, a nonnegative, sparse, and collaborative representation (NSCR) is used to represent the feature vector and solve the coding coefficient matrix; fourthly, the residual similarity is used for classification. The experimental results show that the DenseNet-NSCR classification is better than the other models, and the various evaluation indexes such as specificity and sensitivity are also high, and the method has better robustness and generalization ability through comparison experiment using AlexNet, GoogleNet, and DenseNet-201 models.