Tensor Discriminant Analysis via Compact Feature Representation for Hyperspectral Images Dimensionality Reduction

Motion trajectory recognition is important for characterizing the moving property of an object. The speed and accuracy of trajectory recognition rely on a compact and discriminative feature representation, and the situations of varying rotation, scaling and translation has to be specially considered. In this paper we propose a novel feature extraction method for trajectories. Firstly a trajectory is represented by a proposed point context, which is a rotation-scale-translation (RST) invariant shape descriptor with a flexible tradeoff between computational complexity and discrimination, yet we prove that it is a complete shape descriptor. Secondly, the shape context is nonlinearly mapped to a subspace by kernel nonparametric discriminant analysis (KNDA) to get a compact feature representation, and thus a trajectory is projected to a single point in a low-dimensional feature space. Experimental results show that, the proposed trajectory feature shows encouraging improvement than state-of-art methods.

Hyperspectral image classification (HSIC) leverages the rich spectral information in hyperspectral images (HSIs) and plays an increasingly important role in the field of remote sensing, yet existing methods often fail to fully exploit multi-scale spatial-spectral dependencies. To address this limitation, this work proposes the multi-kernel aggregation network (MKANet) for HSIC. MKANet incorporates a multi-kernel aggregation (MKA) module to capture spatial-spectral features at different scales, enhancing local feature representation and improving the model's adaptability to diverse spatial structures. In addition, a contextual cohesion attention (CCA) module is introduced to model long-range spatial-spectral relationships, strengthening contextual awareness and feature discrimination. The joint design of MKA and CCA modules enables the network to effectively integrate local and global information for more comprehensive feature learning. Experimental results on two publicly available HSI datasets validate the effectiveness of our proposed method.

Hyperspectral images (HSI) have high-dimensional and complex spectral characteristics, with dozens or even hundreds of bands covering the same area of pixels. The rich information of the ground objects makes hyperspectral images widely used in satellite remote sensing. Due to the limitations of remote sensing satellite sensors, hyperspectral images suffer from insufficient spatial resolution. Therefore, utilizing software algorithms to improve the spatial resolution of hyperspectral images has become an urgent problem that needs to be solved. The spatial information and spectral information of hyperspectral images are strongly correlated. If only the spatial resolution is improved, it often damages the spectral information. Inspired by the high correlation between spectral information in adjacent spectral bands of hyperspectral images, a hybrid convolution and spectral symmetry preservation network has been proposed for hyperspectral super-resolution reconstruction. This includes a model to integrate information from neighboring spectral bands to supplement target band feature information. The proposed model introduces flexible spatial-spectral symmetric 3D convolution in the network structure to extract low-resolution and neighboring band features. At the same time, a combination of deformable convolution and attention mechanisms is used to extract information from low-resolution bands. Finally, multiple bands are fused in the reconstruction module, and the high-resolution hyperspectral image containing global information is obtained by Fourier transform upsampling. Experiments were conducted on the indoor hyperspectral image dataset CAVE, the airborne hyperspectral dataset Pavia Center, and Chikusei. In the X2 super-resolution task, the PSNR values achieved on the CAVE, Pavia Center, and Chikusei datasets were 46.335, 36.321, and 46.310, respectively. In the X4 super-resolution task, the PSNR values achieved on the CAVE, Pavia Center, and Chikusei datasets were 41.218, 30.377, and 38.365, respectively. The results show that our method outperforms many advanced algorithms in objective indicators such as PSNR and SSIM while maintaining the spectral characteristics of hyperspectral images.

Creating accurate land use and land cover maps using remote sensing images is one of the most important applications of remotely sensed data. Abundant spectral information in hyperspectral images (HSI) makes it possible to distinguish materials that would not be distinguishable by multi-spectral sensors. Spectral and spatial information from HSI is of primary importance for image classification. In this study, a hybrid stacked autoencoder (SAE) architecture and support vector machine (SVM) classifier was constructed to classify the HSI. The SAE architecture is constituted by stacking a multiple autoencoder (AE) deep learning network that consists in the encoder and decoder process. Spatial features in a neighbor region extracted from the principal component analysis (PCA) and the texture feature extracted from the gray-level cooccurrence matrix (GLCM) were fed into the classifier. It was found that the best result was from the combination of GLCM texture feature, PCA spatial feature, and spectral feature. Meanwhile, the representative features derived from SAE deep learning network were better than the original features. It reminded us that extracting the representative features from hyperspectral images is a key step of improving classification accuracy.

Compact feature representation of person image is important for person re-identification (Re-ID) task. Recently, part-based representation models have been widely studied for extracting the more compact and robust feature representation for person image to improve person Re-ID results. However, existing part-based representation models mostly extract the features of different parts independently which ignore the spatial relationship information among different parts. To address this issue, in this paper we propose a novel deep learning framework, named Part-based Hierarchical Graph Convolutional Network (PH-GCN) for person Re-ID problem. Given a person image, PH-GCN first constructs a hierarchical graph to represent the spatial relationships among different parts. Then, both local and global feature learning is achieved by the feature information passing in PH-GCN, which takes the information of other parts into account for part feature representation. Finally, a perceptron layer is adopted for the final person part label prediction and re-identification. The proposed framework provides a general solution that integrates <i>local</i>, <i>global</i> and <i>structural</i> feature learning simultaneously in a unified end-to-end network representation and learning. Extensive experiments on several widely used benchmark datasets demonstrate the effectiveness and benefits of the proposed PH-GCN approach for person Re-ID task.

Hyperspectral imaging has drawn significant attention in recent years, and its application to object detection and classification is currently an important research topic. However, finding a method to accurately identify objects that only occupy a very small part of an image area remains to be a challenge. In this paper, a novel two-stage deep learning-based hyperspectral neural network (2SHyperNet) suitable for human detection from the sea surface is proposed. The method combines spatial and spectral information of hyperspectral images. Pixel-wise spectral information is used in the first stage to obtain first-stage classification results, and then the results are combined with spatial information to help eliminate unlikely regions, thus, improving the detection accuracy. The method is tested on a data set of real-world airborne hyperspectral images, and its performance is compared with those of several conventional methods. The results show that the proposed method outperforms current state-of-the-art methods.

This work was supported in part by the National Science Fund for Excellent Young Scholars under Grant 61522107, by the Natural Science Foundation of China under Grant 61371180, by the China Aerospace Science and Technology Corporation–Harbin Institute of Technology Joint Center for Technology Innovation Fund under Grant CASC-HIT15-1C03, and by the Smart Sea Technical Innovation Foundation, CSSC-SERI.

Multiview data, characterized by rich features, are crucial in many machine learning applications. However, effectively extracting intraview features and integrating interview information present significant challenges in multiview learning (MVL). Traditional deep network-based approaches often involve learning multiple layers to derive latent. In these methods, the features of different classes are typically implicitly embedded rather than systematically organized. This lack of structure makes it challenging to explicitly map classes to independent principal subspaces in the feature space, potentially causing class overlap and confusion. Consequently, the capability of these representations to accurately capture the intrinsic structure of the data remains uncertain. In this article, we introduce an innovative multiview representation learning (MVRL) by maximizing two information-theoretic metrics: intraview coding rate reduction and interview mutual information. Specifically, in the intraview representation learning, we aim to optimize feature representations by maximizing the coding rate difference between the entire dataset and individual classes. This process expands the feature representation space while compressing the representations within each class, resulting in more compact feature representations within each viewpoint. Subsequently, we align and fuse these view-specific features through space transformation and cross-sample fusion to achieve consistent representation across multiple views. Finally, we maximize information transmission to maintain consistency and correlation among data representations across views. By maximizing mutual information between the consensus representations and view-specific representations, our method ensures that the learned representations capture more concise intrinsic features and correlations among different views, thereby enhancing the performance and generalization ability of MVL. Experiments show that the proposed methods have achieved excellent performance.

It is common practice to use pretrained image recognition models to compute feature representations for visual data. The size of the feature representations can have a noticeable impact on the complexity of the models that use these representations, and by extension on their deployablity and scalability. Therefore it would be beneficial to have compact visual representations that carry as much information as their high-dimensional counterparts. To this end we propose a technique that shrinks a layer by an iterative process in which neurons are removed from the and network is fine tuned. Using this technique we are able to remove 99% of the neurons from the penultimate layer of AlexNet and VGG16, while suffering less than 5% drop in accuracy on CIFAR10, Caltech101 and Caltech256. We also show that our method can reduce the size of AlexNet by 95% while only suffering a 4% reduction in accuracy on Caltech101.

Using unlabeled data in a sparse-coding framework for human activity recognition

The classification methods based on minimum spanning forest (MSF) have yielded impressive results for hyperspectral image. However, previous methods exist several drawbacks, i.e., marker selection methods are easily affected by boundary noise pixels, dissimilarity measure methods between pixels are inaccurate, and also image segmentation process is not robust, since they have not effectively utilized spatial information. To this end, in this paper, novel gradient-based marker selection technique, dissimilarity measures, and adaptive connection weighting method are proposed by making full use of spatial information in hyperspectral image. Concretely, for a given hyperspectral image, a pixel-wise classification is firstly performed, and meanwhile the gradient map is generated by a morphology-based algorithm. Secondly, the most reliable pixels are selected as the markers from the classification map, and then the boundary noise pixels are excluded from the marker map by using the gradient map. Thirdly, several new dissimilarity measures are proposed by incorporating gradient information or probability information of pixels. Furthermore, in the growth procedure of MSF, the connection weighting between pixels is adjusted adaptively to improve the robustness of the MSF algorithm. Finally, when building the final classification map by using the majority voting rule, the labels of the training samples are used to dominate the label prediction. Experimental results are performed on two hyperspectral image sets Indian Pines and University of Pavia with different resolutions and contexts. The proposed approach yields higher classification accuracies compared to previously proposed classification methods, and provides accurate segmentation maps.

Change detection methods using hyperspectral remote sensing can precisely identify differences of the same area at different observing times. However, due to massive spectral bands, current change detection methods are vulnerable to unrelatedspectral and spatial information in hyperspectral images with the stagewise calculation of attention maps. Besides, current change methods arrange hidden change features in a random distribution form, which cannot express a class-oriented discrimination in advance. Moreover, existent deep change methods have not fully considered the hierarchical features’ reuse and the fusion of the encoder–decoder framework. To better handle the mentioned existent problems, the parallel spectral–spatial attention network with feature redistribution loss (TFR-PS2ANet) is proposed. The contributions of this article are summarized as follows: (1) a parallel spectral–spatial attention module (PS2A) is introduced to enhance relevant information and suppress irrelevant information in parallel using spectral and spatial attention maps extracted from the original hyperspectral image patches; (2) the feature redistribution loss function (FRL) is introduced to construct the class-oriented feature distribution, which organizes the change features in advance and improves the discriminative abilities; (3) a two-branch encoder–decoder framework is developed to optimize the hierarchical transfer and change features’ fusion; Extensive experiments were carried out on several real datasets. The results show that the proposed PS2A can enhance significant information effectively and the FRL can optimize the class-oriented feature distribution. The proposed method outperforms most existent change detection methods.

Different types of methods have been developed to retrieve vegetation attributes from remote sensing data, including conventional empirical regressions (i.e., linear regression (LR)), advanced empirical regressions (e.g., multivariable linear regression (MLR), partial least square regression (PLSR)), machine learning (e.g., random forest regression (RFR), decision tree regression (DTR)), and radiative transfer modelling (RTM, e.g., PROSAIL). Given that each algorithm has its own strengths and weaknesses, it is essential to compare them and evaluate their effectiveness. Previous studies have mainly used single-date multispectral imagery or ground-based hyperspectral reflectance data for evaluating the models, while multi-seasonal hyperspectral images have been rarely used. Extensive spectral and spatial information in hyperspectral images, as well as temporal variations of landscapes, potentially influence the model performance. In this research, LR, PLSR, RFR, and PROSAIL, representing different types of methods, were evaluated for estimating vegetation chlorophyll content from bi-seasonal hyperspectral images (i.e., a middle- and a late-growing season image, respectively). Results show that the PLSR and RFR generally performed better than LR and PROSAIL. RFR achieved the highest accuracy for both images. This research provides insights on the effectiveness of different models for estimating vegetation chlorophyll content using hyperspectral images, aiming to support future vegetation monitoring research.

Sparsity-preserving graph construction is investigated for the dimensionality reduction of hyperspectral imagery. In particular, a sparse graph-based discriminant analysis is proposed when labeled samples are available. By forcing the projection to be along the direction where a sample is clustered with within-class samples that best represented it, the discriminative power can be enhanced. The proposed method has no requirement on the number of labeled samples as in traditional linear discriminant analysis, and it can be solved by a simple generalized eigenproblem. The quality of the dimensionality reduction is evaluated by a support vector machine with a composite spatial-spectral kernel. Experimental results demonstrate that the proposed sparse graph-based discriminant analysis can yield superior classification performance with much lower dimensionality as compared to performance on the original data or on data transformed with other dimensionality-reduction approaches.

Compact and discriminative representation of Bag-of-Features