Convolutional neural network (CNN) has demonstrated impressive ability to represent hyperspectral images and to achieve promising results in hyperspectral image classification. However, traditional ...CNN models can only operate convolution on regular square image regions with fixed size and weights, and thus, they cannot universally adapt to the distinct local regions with various object distributions and geometric appearances. Therefore, their classification performances are still to be improved, especially in class boundaries. To alleviate this shortcoming, we consider employing the recently proposed graph convolutional network (GCN) for hyperspectral image classification, as it can conduct the convolution on arbitrarily structured non-Euclidean data and is applicable to the irregular image regions represented by graph topological information. Different from the commonly used GCN models that work on a fixed graph, we enable the graph to be dynamically updated along with the graph convolution process so that these two steps can be benefited from each other to gradually produce the discriminative embedded features as well as a refined graph. Moreover, to comprehensively deploy the multiscale information inherited by hyperspectral images, we establish multiple input graphs with different neighborhood scales to extensively exploit the diversified spectral-spatial correlations at multiple scales. Therefore, our method is termed multiscale dynamic GCN (MDGCN). The experimental results on three typical benchmark data sets firmly demonstrate the superiority of the proposed MDGCN to other state-of-the-art methods in both qualitative and quantitative aspects.
Synthetic aperture radar (SAR) automatic target recognition (ATR) plays an important role in SAR image interpretation. However, at least hundreds of training samples are usually required for each ...target type in the existing SAR ATR algorithms. In this article, a novel few-shot learning framework named hybrid inference network (HIN) is proposed to tackle the problem of SAR target recognition with only a few training samples. The recognition procedure of HIN consists of two main stages. In the first stage, an embedding network is utilized to map the SAR images into an embedding space. In the second stage, a hybrid inference strategy that combines the inductive inference and the transductive inference is adopted to classify the samples in the embedding space. In the inductive inference section, each sample is recognized independently according to a metric based on Euclidean distance. In the transductive inference section, all samples are recognized as a whole according to their manifold structures by label propagation. Finally, in the hybrid inference section, the classification result is obtained by combining the above two inference methods. To train the framework more effectively, a novel loss function named enhanced hybrid loss is proposed to constrain samples to gain better interclass separability in the embedding space. Experimental results on the moving and stationary target acquisition and recognition (MSTAR) benchmark data set illustrate that HIN performs well in few-shot SAR image classification.
Artificial Intelligence (AI) has been widely utilized in automated journalism writing and broadcasting in recent years. However, few systematic studies have been conducted on the differences in brain ...activation between human and artificial voices in newscasts. This paper aims to investigate the psychophysiological effects of the media in Chinese contexts when different agents (AI or human) broadcast different types (emotional/neutral) of news. Comparing the electrophysiological data of the participants' EEG while listening to different newscast agents revealed that brain activity responses were greater when listening to a human newscast than to an AI voice newscast. And β bands in left posterior temporal lobe were significantly different, suggesting that participants' brain were better at processing, comprehending auditory information, and storing working memory when listening to a human reporter than when listening to a voice generated by AI. Moreover, the ERP results and the interaction effect of content valence and agent voice demonstrated that the human voice generated greater cognitive effect, which may reflect participants' trust in the credibility and fluency of the human voice news. This study demonstrates the importance of further research into cognitive effects of AI journalism.
For seeking high enantiopurity, the previously reported thermal asymmetric catalysis is usually carried out at low temperature sometimes with limited yield, that is, the high enantiomeric excess (ee) ...usually at the cost of high yield. Thus, the achieving both high stereoselectivity and yield is an enormous challenge. We report herein two metal nanoparticle (M NP)-loaded and porphyrin-containing homochiral covalent organic framework (CCOF)-based composite catalysts, and their application in the thermally-driven asymmetric one-pot Henry and A
-coupling reactions. All the reactions are conducted at elevated temperatures with both excellent stereoselectivity and yield which resulted from the synergy of CCOF confinement effect and M NP catalytic activation. Notably, the needed thermal energy for the asymmetric reactions herein is derived from the photothermal conversion via porphyrin-based CCOF upon irradiation with visible light. Remarkably, the CCOF confinement effect can be effectively maintained up to 100 °C for the asymmetric one-pot Henry and A
-coupling reactions herein.
Semi-supervised image classification aims to classify a large quantity of unlabeled images by typically harnessing scarce labeled images. Existing semi-supervised methods often suffer from inadequate ...classification accuracy when encountering difficult yet critical images, such as outliers, because they treat all unlabeled images equally and conduct classifications in an imperfectly ordered sequence. In this paper, we employ the curriculum learning methodology by investigating the difficulty of classifying every unlabeled image. The reliability and the discriminability of these unlabeled images are particularly investigated for evaluating their difficulty. As a result, an optimized image sequence is generated during the iterative propagations, and the unlabeled images are logically classified from simple to difficult. Furthermore, since images are usually characterized by multiple visual feature descriptors, we associate each kind of features with a teacher, and design a multi-modal curriculum learning (MMCL) strategy to integrate the information from different feature modalities. In each propagation, each teacher analyzes the difficulties of the currently unlabeled images from its own modality viewpoint. A consensus is subsequently reached among all the teachers, determining the currently simplest images (i.e., a curriculum), which are to be reliably classified by the multi-modal learner. This well-organized propagation process leveraging multiple teachers and one learner enables our MMCL to outperform five state-of-the-art methods on eight popular image data sets.
This article explores the Chinese Xinhai revolutionaries' views of the Young Turk Revolution in 1908, and the Young Turks' semi-official organ Tanin's coverage of the Chinese Xinhai Revolution in ...1911. The Chinese and Ottoman revolutionaries were greatly interested in each other. And there was a striking similarity between their views of each other: both the Chinese and Ottoman revolutionaries analyzed each other's revolution in the context of international politics. They knew a revolution in the other country would change that country's position in international politics and impact the great powers' thinking and behavior, and the response of the great powers in turn would have an influence on the fate of their own country. For Chinese revolutionaries, the Young Turk Revolution proved the need for a revolution in China. The revolutionaries' interests in each other and similarity in their views proved an important but neglected intellectual connection between revolutions in the early twentieth century.
Neuroregeneration in the CNS has proven to be difficult despite decades of research. The old dogma that CNS neurons cannot be regenerated in the adult mammalian brain has been overturned; however, ...endogenous adult neurogenesis appears to be insufficient for brain repair. Stem cell therapy once held promise for generating large quantities of neurons in the CNS, but immunorejection and long-term functional integration remain major hurdles. In this Perspective, we discuss the use of in vivo reprogramming as an emerging technology to regenerate functional neurons from endogenous glial cells inside the brain and spinal cord. Besides the CNS, in vivo reprogramming has been demonstrated successfully in the pancreas, heart, and liver and may be adopted in other organs. Although challenges remain for translating this technology into clinical therapies, we anticipate that in vivo reprogramming may revolutionize regenerative medicine by using a patient’s own internal cells for tissue repair.
In this Perspective, Li and Chen discuss the potential for in vivo reprogramming technology to be used for neural repair in the brain and spinal cord through converting endogenous glial cells into functional neurons.
Abstract
Background
Post-traumatic stress disorder (PTSD) sufferers show problematic patterns of Internet use such as fear of missing out (FOMO) and sharing misinformation and fake news. This study ...aimed to investigate these associations in survivors of the 2008 earthquake in Wenchuan, China.
Methods
A self-reported survey was completed by 356 survivors of the Wenchuan earthquake. A mediated structural equation model was constructed to test a proposed pattern of associations with FOMO as a mediator of the relationship between PTSD symptoms and belief in fake news, as well as moderators of this pathway.
Results
PTSD was directly associated with believing fake news (β = 0.444, p < .001) and with FOMO (β = 0.347, p < .001). FOMO mediated the association between PTSD and fake news belief (β = 0.373, p < .001). Age moderated the direct (β = 0.148, t = 3.097, p = .002) and indirect (β = 0.145, t = 3.122, p = .002) pathways, with effects more pronounced with increasing age. Gender was also a moderator, with the indirect effect present in females but not in males (β = 0.281, t = 6.737, p < .001).
Conclusion
Those with higher PTSD symptoms are more likely to believe fake news and this is partly explained by FOMO. This effect is present in females and not males and is stronger in older people. Findings extend knowledge of the role of psychological variables in problematic Internet use among those with PTSD.
In hyperspectral image (HSI) classification, spatial context has demonstrated its significance in achieving promising performance. However, conventional spatial context-based methods simply assume ...that spatially neighboring pixels should correspond to the same land-cover class, so they often fail to correctly discover the contextual relations among pixels in complex situations, and thus leading to imperfect classification results on some irregular or inhomogeneous regions such as class boundaries. To address this deficiency, we develop a new HSI classification method based on the recently proposed graph convolutional network (GCN), as it can flexibly encode the relations among arbitrarily structured non-Euclidean data. Different from traditional GCN, there are two novel strategies adopted by our method to further exploit the contextual relations for accurate HSI classification. First, since the receptive field of traditional GCN is often limited to fairly small neighborhood, we proposed to capture long-range contextual relations in HSI by performing successive graph convolutions on a learned region-induced graph which is transformed from the original 2-D image grids. Second, we refine the graph edge weight and the connective relationships among image regions simultaneously by learning the improved similarity measurement and the "edge filter," so that the graph can be gradually refined to adapt to the representations generated by each graph convolutional layer. Such updated graph will in turn result in faithful region representations, and vice versa. The experiments carried out on four real-world benchmark data sets demonstrate the effectiveness of the proposed method.
Semantic segmentation (SS) partitions an image into several coherent semantically meaningful parts and classifies each part into one of the pre-determined classes. In this paper, we argue that the ...existing SS methods cannot be reliably applied to autonomous driving system as they ignore the different importance levels of distinct classes for safe driving. For example, pedestrian, car, and bicyclist in the scene are much more important than sky and building when driving a car, so their segmentations should be as accurate as possible. To incorporate the importance information possessed by various object classes, this paper designs an "importance-aware loss" (IAL) that specifically emphasizes the critical objects for autonomous driving. The IAL operates under a hierarchical structure and the classes with different importance are located in different levels so that they are assigned distinct weights. Furthermore, we derive the forward and backward propagation rules for IAL and apply them to four typical deep neural networks for realizing SS in an intelligent driving system. The experiments on CamVid and Cityscapes data sets reveal that, by employing the proposed loss function, the existing deep learning models, including FCN, SegNet, ENet, and ERFNet, are able to consistently obtain the improved segmentation results on the pre-defined important classes for safe driving.