scholarly journals Vehicle Type Recognition Combining Global and Local Features via Two-Stage Classification

2017 ◽  
Vol 2017 ◽  
pp. 1-14 ◽  
Author(s):  
Wei Sun ◽  
Xiaorui Zhang ◽  
Shunshun Shi ◽  
Jun He ◽  
Yan Jin
Keyword(s):  
Nearest Neighbor ◽  
Local Features ◽  
Detection Algorithm ◽  
Local Feature ◽  
Gabor Wavelet ◽  
Global Feature ◽  
Two Stage ◽  
Vehicle Type ◽  
Stage Classification ◽  
Global And Local

This study proposes a new vehicle type recognition method that combines global and local features via a two-stage classification. To extract the continuous and complete global feature, an improved Canny edge detection algorithm with smooth filtering and non-maxima suppression abilities is proposed. To extract the local feature from four partitioned key patches, a set of Gabor wavelet kernels with five scales and eight orientations is introduced. Different from the single-stage classification, where all features are incorporated into one classifier simultaneously, the proposed two-stage classification strategy leverages two types of features and classifiers. In the first stage, the preliminary recognition of large vehicle or small vehicle is conducted based on the global feature via a k-nearest neighbor probability classifier. Based on the preliminary result, the specific recognition of bus, truck, van, or sedan is achieved based on the local feature via a discriminative sparse representation based classifier. We experiment with the proposed method on the public and established datasets involving various challenging cases, such as partial occlusion, poor illumination, and scale variation. Experimental results show that the proposed method outperforms existing state-of-the-art methods.

scholarly journals Memory-Augmented Transformer for Remote Sensing Image Semantic Segmentation

2021 ◽  
Vol 13 (22) ◽  
pp. 4518
Author(s):  
Xin Zhao ◽  
Jiayi Guo ◽  
Yueting Zhang ◽  
Yirong Wu
Keyword(s):  
Remote Sensing ◽  
Feature Extraction ◽  
Semantic Segmentation ◽  
Local Features ◽  
Local Feature ◽  
Global Information ◽  
Deep Convolutional Neural Networks ◽  
Global Representation ◽  
Local Feature Extraction ◽  
Global And Local

The semantic segmentation of remote sensing images requires distinguishing local regions of different classes and exploiting a uniform global representation of the same-class instances. Such requirements make it necessary for the segmentation methods to extract discriminative local features between different classes and to explore representative features for all instances of a given class. While common deep convolutional neural networks (DCNNs) can effectively focus on local features, they are limited by their receptive field to obtain consistent global information. In this paper, we propose a memory-augmented transformer (MAT) to effectively model both the local and global information. The feature extraction pipeline of the MAT is split into a memory-based global relationship guidance module and a local feature extraction module. The local feature extraction module mainly consists of a transformer, which is used to extract features from the input images. The global relationship guidance module maintains a memory bank for the consistent encoding of the global information. Global guidance is performed by memory interaction. Bidirectional information flow between the global and local branches is conducted by a memory-query module, as well as a memory-update module, respectively. Experiment results on the ISPRS Potsdam and ISPRS Vaihingen datasets demonstrated that our method can perform competitively with state-of-the-art methods.

Improved Elastic Model Matching Algorithm for Facial Expression Recognition

2013 ◽  
Vol 397-400 ◽  
pp. 2143-2147
Author(s):  
Wen Dong Zhao ◽  
Jie Zhang ◽  
You Dong Zhang
Keyword(s):  
Nearest Neighbor ◽  
Detection Algorithm ◽  
Gabor Wavelet ◽  
Similar Degree ◽  
Elastic Model ◽  
Expression Recognition ◽  
Model Matching ◽  
K Nearest Neighbor ◽  
Matching Algorithm ◽  
Expression Template

Key frames detected in Video stream contain sufficient expression information. In order to classify and recognize these expression information, a new elastic model matching algorithm is proposed in this paper. Firstly, expression template is transformed by Gabor wavelet , and the detection algorithm of the key expression in the template image is used. According to the feature information of the key expression, structure expression elastic graph, then by changing the key location in the expression template graph and doing non-rigid match of the expression template and expression elastic graph which is measured, so that similar degree between them is got. Finally, by improving the K-nearest neighbor classification strategy, the effective classification and recognition of measured image expression is achieved.

Robust two-stage face recognition approach using global and local features

2011 ◽  
Vol 28 (11) ◽  
pp. 1085-1098 ◽  
Author(s):  
Chandan Singh ◽  
Ekta Walia ◽  
Neerja Mittal
Keyword(s):  
Face Recognition ◽  
Local Features ◽  
Two Stage ◽  
Global And Local

scholarly journals FastText-Based Local Feature Visualization Algorithm for Merged Image-Based Malware Classification Framework for Cyber Security and Cyber Defense

Mathematics ◽  
2020 ◽  
Vol 8 (3) ◽  
pp. 460 ◽  
Author(s):  
Sejun Jang ◽  
Shuyu Li ◽  
Yunsick Sung
Keyword(s):  
Cyber Security ◽  
Local Features ◽  
Cyber Attacks ◽  
Local Feature ◽  
Easy Access ◽  
Malware Classification ◽  
Classification Framework ◽  
Starting Point ◽  
Global And Local ◽  
Feature Visualization

The importance of cybersecurity has recently been increasing. A malware coder writes malware into normal executable files. A computer is more likely to be infected by malware when users have easy access to various executables. Malware is considered as the starting point for cyber-attacks; thus, the timely detection, classification and blocking of malware are important. Malware visualization is a method for detecting or classifying malware. A global image is visualized through binaries extracted from malware. The overall structure and behavior of malware are considered when global images are utilized. However, the visualization of obfuscated malware is tough, owing to the difficulties encountered when extracting local features. This paper proposes a merged image-based malware classification framework that includes local feature visualization, global image-based local feature visualization, and global and local image merging methods. This study introduces a fastText-based local feature visualization method: First, local features such as opcodes and API function names are extracted from the malware; second, important local features in each malware family are selected via the term frequency inverse document frequency algorithm; third, the fastText model embeds the selected local features; finally, the embedded local features are visualized through a normalization process. Malware classification based on the proposed method using the Microsoft Malware Classification Challenge dataset was experimentally verified. The accuracy of the proposed method was approximately 99.65%, which is 2.18% higher than that of another contemporary global image-based approach.

scholarly journals A Geometrical-Information-Assisted Approach for Local Feature Matching

2019 ◽  
Vol 2019 ◽  
pp. 1-15
Author(s):  
Buhai Shi ◽  
Qingming Zhang ◽  
Haibo Xu
Keyword(s):  
Nearest Neighbor ◽  
Feature Matching ◽  
Bayes Theorem ◽  
Local Features ◽  
Local Feature ◽  
Feature Points ◽  
New Approach ◽  
Feature Descriptors ◽  
Geometrical Information

This paper presents a geometrical-information-assisted approach for matching local features. With the aid of Bayes’ theorem, it is found that the posterior confidence of matched features can be improved by introducing global geometrical information given by distances between feature points. Based on this result, we work out an approach to obtain the geometrical information and apply it to assist matching features. The pivotal techniques in this paper include (1) exploiting elliptic parameters of feature descriptors to estimate transformations that map feature points in images to points in an assumed plane; (2) projecting feature points to the assumed plane and finding a reliable referential point in it; (3) computing differences of the distances between the projected points and the referential point. Our new approach employs these differences to assist matching features, reaching better performance than the nearest neighbor-based approach in precision versus the number of matched features.

Object Representation Fusing Global and Local Features

2013 ◽  
Vol 373-375 ◽  
pp. 1022-1026
Author(s):  
Tian Wen Li ◽  
Yun Gao
Keyword(s):  
Object Tracking ◽  
Object Representation ◽  
Feature Fusion ◽  
Local Features ◽  
Partial Occlusion ◽  
Global Feature ◽  
Complex Environments ◽  
Complex Scenes ◽  
Global Representation ◽  
Global And Local

In the actual complex scenes, multi-feature fusion has become a valid method of object representation for tracking video motion targets. Two keys about multi-feature fusion are how to select some valid features and how to fuse the features. In this paper, we propose an object representation fusing global and local features for object tracking. In our method, we select a common hue histogram as the global feature and use a valid SIFT feature as the local feature. In the tracking frame of particle filter, the tracking results show that our proposed object representation can better restrain the disturbing of complex environments with abrupt illumination and partial occlusion, than color-based global representation.

Research of Face Recognition System Based on TMS320DM642

2013 ◽  
Vol 341-342 ◽  
pp. 975-979
Author(s):  
Zhen Liu ◽  
Ji Chao Yuan ◽  
Shuai Mei ◽  
Xiong Shi
Keyword(s):  
Face Recognition ◽  
Digital Signal Processor ◽  
Skin Color ◽  
Nearest Neighbor ◽  
Principal Component ◽  
Recognition System ◽  
Detection Algorithm ◽  
Gabor Wavelet ◽  
Face Recognition System ◽  
System Solution

This paper proposes an embedded face recognition system solution. The core of hardware architecture of the system is TMS320DM642 digital signal processor (DSP). The face recognition algorithm of the system mainly comprises improved face detection algorithm which is based on skin color, Gabor wavelet feature extraction algorithm, Principal Component Analysis (PCA) algorithm and nearest neighbor classifier algorithm. The results of testing on recognition efficiency and execution time show that the system can work stably and realize face recognition quickly and accurately.

Sign in / Sign up

Export Citation Format

Share Document