A Novel HOG Descriptor with Spatial Multi-Scale Feature for FER

HOG Feature is an efficient edge information descriptor, but it ignores the spatial arrangement of local FER features. In this respect, this paper puts forward a spatial multi-scale model based on an improved HOG algorithm which uses canny operator instead of traditional gradient operator. After the image is divided into a series of sub-regions layer by layer, the histogram of orient gradients for each sub-region is calculated and connected in sequence to obtain the spatial multi-scale HOG feature of whole image. Compared with traditional HOG and the improved PHOG, the proposed SMS_HOG algorithm acquires 5% recognition rate improvement and 50% processing time reduction.

Download Full-text

A Comparative Study on the Method of Extracting Edge and Contour Information of Multifunctional Digital Ship Image

Polish Maritime Research ◽

10.1515/pomr-2017-0127 ◽

2017 ◽

Vol 24 (s3) ◽

pp. 228-234 ◽

Cited By ~ 1

Author(s):

Fangping Yin

Keyword(s):

Image Quality ◽

Comparative Study ◽

Digital Image ◽

The Other ◽

Scale Morphology ◽

Canny Operator ◽

Edge Information ◽

Multi Scale ◽

Contour Information ◽

The Common

Abstract The result of the extraction of the edge and contour information of the multifunctional digital ship image directly affects the evaluation and recognition of the subsequent image quality. At present, the common method used to extract the edge contour information is based on the Canny operator, and there is a problem that the edge is not clear.In order to obtain more accurate edge information, a method of extracting edge and contour information of multimedia digital image based on multi-scale morphology is proposed. Firstly, the digital ship image is made double filter and the fuzzy threshold segmentation, and then the edge and contour information is extracted by multi-scale morphology. Experiments show that the proposed method can obtain more accurate edge information compared with the other methods.

Download Full-text

Multi-Level and Multi-Scale Feature Aggregation Network for Semantic Segmentation in Vehicle-Mounted Scenes

Sensors ◽

10.3390/s21093270 ◽

2021 ◽

Vol 21 (9) ◽

pp. 3270

Author(s):

Yong Liao ◽

Qiong Liu

Keyword(s):

Receptive Fields ◽

Semantic Segmentation ◽

Layer By Layer ◽

Scale Feature ◽

Backbone Networks ◽

Multi Scale ◽

Backbone Network ◽

Feature Aggregation ◽

Multi Level ◽

High Level

The main challenges of semantic segmentation in vehicle-mounted scenes are object scale variation and trading off model accuracy and efficiency. Lightweight backbone networks for semantic segmentation usually extract single-scale features layer-by-layer only by using a fixed receptive field. Most modern real-time semantic segmentation networks heavily compromise spatial details when encoding semantics, and sacrifice accuracy for speed. Many improving strategies adopt dilated convolution and add a sub-network, in which either intensive computation or redundant parameters are brought. We propose a multi-level and multi-scale feature aggregation network (MMFANet). A spatial pyramid module is designed by cascading dilated convolutions with different receptive fields to extract multi-scale features layer-by-layer. Subseqently, a lightweight backbone network is built by reducing the feature channel capacity of the module. To improve the accuracy of our network, we design two additional modules to separately capture spatial details and high-level semantics from the backbone network without significantly increasing the computation cost. Comprehensive experimental results show that our model achieves 79.3% MIoU on the Cityscapes test dataset at a speed of 58.5 FPS, and it is more accurate than SwiftNet (75.5% MIoU). Furthermore, the number of parameters of our model is at least 53.38% less than that of other state-of-the-art models.

Download Full-text

Three-dimensional Reconstruction of Microscopic Image Based on Multi-ST Algorithm

Journal of Imaging Science and Technology ◽

10.2352/j.imagingsci.technol.2020.64.2.020506 ◽

2020 ◽

Vol 64 (2) ◽

pp. 20506-1-20506-7

Author(s):

Min Zhu ◽

Rongfu Zhang ◽

Pei Ma ◽

Xuedian Zhang ◽

Qi Guo

Keyword(s):

3D Reconstruction ◽

Three Dimensional ◽

Scale Model ◽

Microscopic Image ◽

Multi Scale ◽

Gaussian Pyramid ◽

Cost Aggregation ◽

Dimensional Reconstruction ◽

Image Pairs ◽

Accurate Matching

Abstract Three-dimensional (3D) reconstruction is extensively used in microscopic applications. Reducing excessive error points and achieving accurate matching of weak texture regions have been the classical challenges for 3D microscopic vision. A Multi-ST algorithm was proposed to improve matching accuracy. The process is performed in two main stages: scaled microscopic images and regularized cost aggregation. First, microscopic image pairs with different scales were extracted according to the Gaussian pyramid criterion. Second, a novel cost aggregation approach based on the regularized multi-scale model was implemented into all scales to obtain the final cost. To evaluate the performances of the proposed Multi-ST algorithm and compare different algorithms, seven groups of images from the Middlebury dataset and four groups of experimental images obtained by a binocular microscopic system were analyzed. Disparity maps and reconstruction maps generated by the proposed approach contained more information and fewer outliers or artifacts. Furthermore, 3D reconstruction of the plug gauges using the Multi-ST algorithm showed that the error was less than 0.025 mm.

Download Full-text

Effects of parametric uncertainty on multi-scale model predictions of shock response of a pressed energetic material

Journal of Applied Physics ◽

10.1063/1.5098955 ◽

2019 ◽

Vol 125 (23) ◽

pp. 235104 ◽

Cited By ~ 6

Author(s):

Sangyup Lee ◽

Oishik Sen ◽

Nirmal Kumar Rai ◽

Nicholas J. Gaul ◽

K. K. Choi ◽

...

Keyword(s):

Energetic Material ◽

Parametric Uncertainty ◽

Scale Model ◽

Multi Scale ◽

Model Predictions ◽

Shock Response

Download Full-text

Multi-Scale Aggregation Graph Neural Networks Based on Feature Similarity for Semi-Supervised Learning

Entropy ◽

10.3390/e23040403 ◽

2021 ◽

Vol 23 (4) ◽

pp. 403

Author(s):

Xun Zhang ◽

Lanyan Yang ◽

Bin Zhang ◽

Ying Liu ◽

Dong Jiang ◽

...

Keyword(s):

Neural Network ◽

Supervised Learning ◽

Structured Data ◽

Similarity Matrix ◽

Scale Feature ◽

Multi Scale ◽

Adjacent Node ◽

Feature Similarity ◽

The Mean ◽

Graph Neural Networks

The problem of extracting meaningful data through graph analysis spans a range of different fields, such as social networks, knowledge graphs, citation networks, the World Wide Web, and so on. As increasingly structured data become available, the importance of being able to effectively mine and learn from such data continues to grow. In this paper, we propose the multi-scale aggregation graph neural network based on feature similarity (MAGN), a novel graph neural network defined in the vertex domain. Our model provides a simple and general semi-supervised learning method for graph-structured data, in which only a very small part of the data is labeled as the training set. We first construct a similarity matrix by calculating the similarity of original features between all adjacent node pairs, and then generate a set of feature extractors utilizing the similarity matrix to perform multi-scale feature propagation on graphs. The output of multi-scale feature propagation is finally aggregated by using the mean-pooling operation. Our method aims to improve the model representation ability via multi-scale neighborhood aggregation based on feature similarity. Extensive experimental evaluation on various open benchmarks shows the competitive performance of our method compared to a variety of popular architectures.

Download Full-text