Wheat head detection can estimate various wheat traits, such as density, health, and the presence of head. However, traditional methods have a huge array problems, including low efficiency, strong subjectivity, poor accuracy. In this paper, method wheat-head based on deep neural network is proposed to enhance speed accuracy detection. The YOLOv4 taken basic network. backbone part in enhanced by adding dual spatial pyramid pooling (SPP) networks improve ability feature learning increase...

Text sentiment tendency analysis is a hot task in natural language processing. And text as the essential expression form of language, both individual word information, and overall utterance, deserves to be focused on. This paper proposes fusion model achieve high precision analysis. combines advantages CNN extract local information BiLSTM contextual association introduces attention mechanism increase focus on words with solid emotional text. The training datasets are comments that crawled...

With the continuous development of artificial intelligence, embedding object detection algorithms into autonomous underwater detectors for marine garbage cleanup has become an emerging application area. Considering complexity environment and low resolution images taken by detectors, this paper proposes improved algorithm based on Mask R-CNN, with aim achieving high accuracy instance segmentation. First, idea dilated convolution is introduced in Feature Pyramid Network to enhance feature...

With the development of Internet technology, online loans continue to enter public eye, individuals and small businesses must access more loan opportunities, it is important for platforms effectively reduce credit crisis associated with customer defaults. This paper uses default dataset from lending club. The ADASYN (Adaptive synthetic sampling approach) method adopted cope class imbalance problem dataset. In order improve prediction accuracy, this utilizes Blending fuse three models:...

Abstract To properly repair and maintain implants, which are bone tissue implants that replace natural tooth roots, it is crucial to accurately identify their brand specification. Deep learning has demonstrated outstanding capabilities in analysis, such as image identification classification, by the inherent rules degrees of representation data models. The purpose this study evaluate deep algorithms supporting application software for ability recognize categorize three dimensional (3D) Cone...

With Yangtze River through the downtown area, large urban river have great influence on climate in Wuhan, which is typical representative of summer hot and winter cold area. Therefore, micro-climate <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">[1] [2]</sup> conditions blocks Wuhan city are research value. In order to acquire realities study regulation law give directions planning architecture design, a block alongside Hankou, city, was chosen...

The processing of unknown words is an important issue that impedes the performance many natural language systems. In terms financial public opinion, due to a wide variety terminology in sector and suddenness opinion events as well proliferation network vocabulary, new emerge while traditional text classification method basically incapable these unrecognized characters. this paper, approach equipped with directional substitution model for proposed. Based on semantic similarity context word2ve...

Head pose estimation from single 2D images has been considered as an important and challenging research task in computer vision. This paper presents a novel head method which utilizes the shape model of Basel face five fiducial points faces. It adjusts deformation according to Laplace distribution afford variation across different persons. A new matching based on PSO (particle swarm optimization) algorithm is applied both reduce time cost reconstruction achieve higher accuracy than...

As a basic and key attribute of human beings, race plays an indispensable role in face analysis. Traditional machine learning methods all tackle the problem classification combination with two separate steps: extracting artificially designed features training proper classifier these features. Some convolutional neural networks have also been proposed to deal this problem, but get unsatisfactory accuracies. In paper, we propose improved deep network based on existing network. The uses branch...

With the development of deep learning technology, handwritten character recognition has become basic hotspot research. At same time, handwriting in minority languages is gradually worthy attention. In this paper, a robust and novel improved convolutional neural network method proposed to recognize normal digital images (Yi language). To improve accuracy fonts, there are four improvements our work. Firstly, separable convolution adopted cope with problem large parameters. Secondly, data...

In the background of big data and better computer hardware, convolutional neural network (CNN) has also made great progress. Since AlexNet model won ILSVRC-2012 competition, CNN been widely applied to version tasks including prediction task face attributes. Different with single attribute prediction, multi-attribute is subject different attributes difficult achieve particularly good performance. With aim building a performance simple framework, ATNet born as based 4 convolution layers. Based...

In this paper, a new fractional Fourier transform-based image fusion method is proposed. All the source images are decomposed by transform, and then order select adaptively. It has well solved poor adaptability of previous methods. The experiments show that algorithm particularly effective not only on general fusion, but also have good effect for infrared visible fusion.

The problem of image-based head pose estimation attracts intensive attention due to a large number applications such as face analysis and modeling. Existing methods often convert into the classification problem, but ignored non-stationary appearance change brought about by equally distributed bins. This paper targets via deep neural decision trees, where non-linear property representative is learned together with bin probability. First, we use Convolutional Neural Network (CNN) get related...

The special climate effects of waterbody have greater influence on the micro-climate waterfront. Therefore, taking a typical block alongside Yangtze River as example, researchers make field measurement at fixed locations and analyze cooling effect river wind. It aims to find out adjustment mechanism River, thus providing guidance for urban planning architecture design.