Cartography includes two major tasks: map making and map application, which is inextricably linked to artificial intelligence technology. The cartographic expert system experienced the intelligent expression of symbolism. After the spatial optimization decision of behaviorism intelligent expression, cartography faces the combination of deep learning under connectionism to improve the intelligent level of cartography. This paper discusses three problems about the proposition of “deep learning + cartography”. One is the consistency between the deep learning method and the map space problem solving strategy, based on gradient descent, local correlation, feature reduction and non-linear nature that answer the feasibility of the combination of “deep learning + cartography”; the second is to analyze the challenges faced by the combination of cartography from its unique disciplinary characteristics and technical environment, involving the non-standard organization of map data, professional requirements for sample establishment, the integration of geometric and geographical features, as well as the inherent spatial scale of the map; thirdly, the entry points and specific methods for integrating map making and map application into deep learning are discussed respectively.

Monitoring marine biodiversity is a challenge in some vulnerable and difficult-to-access habitats, such as underwater caves. Underwater caves are a great focus of biodiversity, concentrating a large number of species in their environment. However, most of the sessile species that live on the rocky walls are very vulnerable, and they are often threatened by different pressures. The use of these spaces as a destination for recreational divers can cause different impacts on the benthic habitat. In this work, we propose a methodology based on video recordings of cave walls and image analysis with deep learning algorithms to estimate the spatial density of structuring species in a study area. We propose a combination of automatic frame overlap detection, estimation of the actual extent of surface cover, and semantic segmentation of the main 10 species of corals and sponges to obtain species density maps. These maps can be the data source for monitoring biodiversity over time. In this paper, we analyzed the performance of three different semantic segmentation algorithms and backbones for this task and found that the Mask R-CNN model with the Xception101 backbone achieves the best accuracy, with an average segmentation accuracy of 82%.

Abrupt changes in environmental temperature, wind and humidity can lead to great threats to human life safety. The Gansu marathon disaster of China highlights the importance of early warning of hypothermia from extremely low apparent temperature (AT). Here a deep convolutional neural network model together with a statistical downscaling framework is developed to forecast environmental factors for 1 to 12 h in advance to evaluate the effectiveness of deep learning for AT prediction at 1 km resolution. The experiments use data for temperature, wind speed and relative humidity in ERA-5 and the results show that the developed deep learning model can predict the upcoming extreme low temperature AT event in the Gansu marathon region several hours in advance with better accuracy than climatological and persistence forecasting methods. The hypothermia time estimated by the deep learning method with a heat loss model agrees well with the observed estimation at 3-hour lead. Therefore, the developed deep learning forecasting method is effective for short-term AT prediction and hypothermia warnings at local areas.

To save patients’ lives, it is important to go for an early diagnosis of intracranial hemorrhage (ICH). For diagnosing ICH, the widely used method is non-contrast computed tomography (NCCT). It has fast acquisition and availability in medical emergency facilities. To predict hematoma progression and mortality, it is important to estimate the volume of intracranial hemorrhage. Radiologists can manually delineate the ICH region to estimate the hematoma volume. This process takes time and undergoes inter-rater variability. In this research paper, we develop and discuss a fine segmentation model and a coarse model for intracranial hemorrhage segmentations. Basically, two different models are discussed for intracranial hemorrhage segmentation. We trained a 2DDensNet in the first model for coarse segmentation and cascaded the coarse segmentation mask output in the fine segmentation model along with input training samples. A nnUNet model is trained in the second fine stage and will use the segmentation labels of the coarse model with true labels for intracranial hemorrhage segmentation. An optimal performance for intracranial hemorrhage segmentation solution is obtained.

Recognizing the discipline category of the abstract text is of great significance for automatic text recommendation and knowledge mining. Therefore, this study obtained the abstract text of social science and natural science in the Web of Science 2010-2020, and used the machine learning model SVM and deep learning model TextCNN and SCI-BERT models constructed a discipline classification model. It was found that the SCI-BERT model had the best performance. The precision, recall, and F1 were 86.54%, 86.89%, and 86.71%, respectively, and the F1 is 6.61% and 4.05% higher than SVM and TextCNN. The construction of this model can effectively identify the discipline categories of abstracts, and provide effective support for automatic indexing of subjects.

In agriculture, crop yield and quality are critical for global food supply and human survival. Challenges such as plant leaf diseases necessitate a fast, automatic, economical, and accurate method. This paper utilizes deep learning, transfer learning, and specific feature learning modules (CBAM, Inception-ResNet) for their outstanding performance in image processing and classification. The ResNet model, pretrained on ImageNet, serves as the cornerstone, with introduced feature learning modules in our IRCResNet model. Experimental results show our model achieves an average prediction accuracy of 96.8574% on public datasets, thoroughly validating our approach and significantly enhancing plant leaf disease identification.