Aerial scene classification, which aims to automatically label an aerial image with a specific semantic category, is a fundamental problem for understanding high-resolution remote sensing imagery. In recent years, it has become an active task in remote sensing area and numerous algorithms have been proposed for this task, including many machine learning and data-driven approaches. However, the existing datasets for aerial scene classification like UC-Merced dataset and WHU-RS19 are with relatively small sizes, and the results on them are already saturated. This largely limits the development of scene classification algorithms. This paper describes the Aerial Image Dataset (AID): a large-scale dataset for aerial scene classification. The goal of AID is to advance the state-of-the-arts in scene classification of remote sensing images. For creating AID, we collect and annotate more than ten thousands aerial scene images. In addition, a comprehensive review of the existing aerial scene classification techniques as well as recent widely-used deep learning methods is given. Finally, we provide a performance analysis of typical aerial scene classification and deep learning approaches on AID, which can be served as the baseline results on this benchmark.
Abstract:Learning efficient image representations is at the core of the scene classification task of remote sensing imagery. The existing methods for solving the scene classification task, based on either feature coding approaches with low-level hand-engineered features or unsupervised feature learning, can only generate mid-level image features with limited representative ability, which essentially prevents them from achieving better performance. Recently, the deep convolutional neural networks (CNNs), which are hierarchical architectures trained on large-scale datasets, have shown astounding performance in object recognition and detection. However, it is still not clear how to use these deep convolutional neural networks for high-resolution remote sensing (HRRS) scene classification. In this paper, we investigate how to transfer features from these successfully pre-trained CNNs for HRRS scene classification. We propose two scenarios for generating image features via extracting CNN features from different layers. In the first scenario, the activation vectors extracted from fully-connected layers are regarded as the final image features; in the second scenario, we extract dense features from the last convolutional layer at multiple scales and then encode the dense features into global image features through commonly used feature coding approaches. Extensive experiments on two public scene classification datasets demonstrate that the image features obtained by the two proposed scenarios, even with a simple linear classifier, can result in remarkable performance and improve the state-of-the-art by a significant margin. The results reveal that the features Remote Sens. 2015, 7 14681 from pre-trained CNNs generalize well to HRRS datasets and are more expressive than the low-and mid-level features. Moreover, we tentatively combine features extracted from different CNN models for better performance.
Iron homeostasis disturbance has been implicated in Alzheimer’s disease (AD), and excess iron exacerbates oxidative damage and cognitive defects. Ferroptosis is a nonapoptotic form of cell death dependent upon intracellular iron. However, the involvement of ferroptosis in the pathogenesis of AD remains elusive. Here, we report that ferroportin1 (Fpn), the only identified mammalian nonheme iron exporter, was downregulated in the brains of APPswe/PS1dE9 mice as an Alzheimer’s mouse model and Alzheimer’s patients. Genetic deletion of Fpn in principal neurons of the neocortex and hippocampus by breeding Fpnfl/fl mice with NEX-Cre mice led to AD-like hippocampal atrophy and memory deficits. Interestingly, the canonical morphological and molecular characteristics of ferroptosis were observed in both Fpnfl/fl/NEXcre and AD mice. Gene set enrichment analysis (GSEA) of ferroptosis-related RNA-seq data showed that the differentially expressed genes were highly enriched in gene sets associated with AD. Furthermore, administration of specific inhibitors of ferroptosis effectively reduced the neuronal death and memory impairments induced by Aβ aggregation in vitro and in vivo. In addition, restoring Fpn ameliorated ferroptosis and memory impairment in APPswe/PS1dE9 mice. Our study demonstrates the critical role of Fpn and ferroptosis in the progression of AD, thus provides promising therapeutic approaches for this disease.
We sought to evaluate the performance of 68 Ga-DOTA-FAPI-04 ( 68 Ga-FAPI) PET/MR for the diagnosis of primary tumor and metastatic lesions in patients with gastric carcinomas and to compare the results with those of 18 F-FDG PET/CT. Methods: Twenty patients with histologically proven gastric carcinomas were recruited, and each patient underwent both 18 F-FDG PET/CT and 68 Ga-FAPI PET/MR. A visual scoring system was established to compare the detectability of primary tumors and metastases in different organs/regions (the peritoneum, abdominal lymph nodes, supradiaphragmatic lymph nodes, liver, ovary, bone, and other tissues). The original maximum standardized uptake value (SUVmax) and normalized SUVmax (calculated by dividing a lesion's original SUVmax with the mean SUV of the descending aorta) of selected lesions on both 18 F-FDG PET/CT and 68 Ga-FAPI PET/MR were measured.Original/normalized SUVmax-FAPI and SUVmax-FDG were compared for patient-based (including a single lesion with the highest activity uptake in each organ/region) and lesion-based (including all lesions [≤ 5] or the 5 lesions with highest activity [> 5]) analyses, respectively. Results:The 20 recruited patients (median age: 56.0 y; range: 29-70 y) included 9 men and 11 women, 14 patients for initial staging and 6 for recurrence detection. 68 Ga-FAPI PET was superior to 18 F-FDG PET for primary tumor detection (100.00% [14/14] vs 71.43% [10/14], p = 0.034), and the former had higher tracer uptake levels (p < 0.05). 68 Ga-FAPI PET was superior to 18 F-FDG PET in both patient-based and lesion-based evaluation except for the metastatic lesions in supradiaphragmatic lymph nodes and ovaries. Additionally, multiple sequences of MR images were beneficial for the interpretation of hepatic metastases in 3 patients, uterine and rectal metastases in 1 patient, ovarian lesions in 7 patients, and osseous metastases in 2 patients. Conclusion:68 Ga-FAPI PET/MR outperformed 18 F-FDG PET/CT in visualizing the primary and most metastatic lesions of gastric cancer, and might be a promising method with the potential of replacing 18 F-FDG PET/CT.
Scene classification plays an important role in the interpretation of remotely sensed high-resolution imagery. However, the performance of scene classification strongly relies on the discriminative power of feature representation, which is generally hand-engineered and requires a huge amount of domain-expert knowledge as well as time-consuming hand tuning. Recently, unsupervised feature learning (UFL) provides an alternative way to automatically learn discriminative feature representation from images. However, the performances achieved by conventional UFL methods are not comparable to the state-ofthe-art, mainly due to the neglect of locally substantial image structures. This paper presents an improved UFL algorithm based on spectral clustering, named UFL-SC, which cannot only adaptively learn good local feature representations but also discover intrinsic structures of local image patches. In contrast to the standard UFL pipeline, UFL-SC first maps the original image patches into a low-dimensional and intrinsic feature space by linear manifold analysis techniques, and then learns a dictionary (e.g., using K-means clustering) on the patch manifold for feature encoding. To generate a feature representation for each local patch, an explicit parameterized feature encoding method, i.e., triangle encoding, is applied with the learned dictionary on the same patch manifold. The holistic feature representation of image scenes is finally obtained by building a bag-of-visual-words (BOW) model of the encoded local features. Experiments demonstrate that the proposed UFL-SC algorithm can extract efficient local features for image scenes and show comparable performance to the state-of-the-art approach on open scene classification benchmark.
Alzheimer's disease (AD) is a complex neurodegenerative disorder characterized in the early stages by loss of learning and memory. However, the mechanism underlying these symptoms remains unclear. The best correlation between cognitive decline and pathological changes is in synaptic dysfunction. Histopathological hallmarks of AD are the abnormal aggregation of Aβ and Tau. Evidence suggests that Aβ and Tau oligomers contribute to synaptic loss in AD. Recently, direct links between epigenetic alterations, such as dysfunction in non-coding RNAs (ncRNAs), and synaptic pathologies have emerged, raising interest in exploring the potential roles of ncRNAs in the synaptic deficits in AD. In this paper, we summarize the potential roles of Aβ, Tau, and epigenetic alterations (especially by ncRNAs) in the synaptic dysfunction of AD and discuss the novel findings in this area.
Aberrant regulation of microRNAs (miRNAs) has been implicated in the pathogenesis of Alzheimer’s disease (AD), but most abnormally expressed miRNAs found in AD are not regulated by synaptic activity. Here we report that dysfunction of miR-135a-5p/Rock2/Add1 results in memory/synaptic disorder in a mouse model of AD. miR-135a-5p levels are significantly reduced in excitatory hippocampal neurons of AD model mice. This decrease is tau dependent and mediated by Foxd3. Inhibition of miR-135a-5p leads to synaptic disorder and memory impairments. Furthermore, excess Rock2 levels caused by loss of miR-135a-5p plays an important role in the synaptic disorder of AD via phosphorylation of Ser726 on adducin 1 (Add1). Blocking the phosphorylation of Ser726 on Add1 with a membrane-permeable peptide effectively rescues the memory impairments in AD mice. Taken together, these findings demonstrate that synaptic-related miR-135a-5p mediates synaptic/memory deficits in AD via the Rock2/Add1 signaling pathway, illuminating a potential therapeutic strategy for AD.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.