Arbitrary Axis-aligned and Multi-scale Drug Recognition System

2020 ◽  
Author(s):  
Che-Wen Chen ◽  
An-Chao Tsai ◽  
Wei-Yen Chang ◽  
Hsuan-Fu Huang ◽  
Jhing-Fa Wang
Keyword(s):  
Recognition System ◽  
Multi Scale ◽  
Arbitrary Axis

scholarly journals MCSM-Wri: A Small-Scale Motion Recognition Method Using WiFi Based on Multi-Scale Convolutional Neural Network

Sensors ◽  
2019 ◽  
Vol 19 (19) ◽  
pp. 4162
Author(s):  
Ma ◽  
Huang ◽  
Li ◽  
Huang ◽  
Ma ◽  
...  
Keyword(s):  
Environmental Changes ◽  
Recognition System ◽  
Letter Recognition ◽  
Small Scale ◽  
Motion Recognition ◽  
Multi Scale ◽  
Scale Characteristics ◽  
Scale Motion ◽  
Device Free ◽  
Fold Cross Validation

environmental perception technology based onWiFi, and some state-of-the-art techniques haveemerged. The wide application of small-scale motion recognition has aroused people’s concern.Handwritten letter is a kind of small scale motion, and the recognition for small-scale motion basedon WiFi has two characteristics. Small-scale action has little impact on WiFi signals changes inthe environment. The writing trajectories of certain uppercase letters are the same as the writingtrajectories of their corresponding lowercase letters, but they are different in size. These characteristicsbring challenges to small-scale motion recognition. The system for recognizing small-scale motion inmultiple classes with high accuracy urgently needs to be studied. Therefore, we propose MCSM-Wri,a device-free handwritten letter recognition system using WiFi, which leverages channel stateinformation (CSI) values extracted from WiFi packets to recognize handwritten letters, includinguppercase letters and lowercase letters. Firstly, we conducted data preproccessing to provide moreabundant information for recognition. Secondly, we proposed a ten-layers convolutional neuralnetwork (CNN) to solve the problem of the poor recognition due to small impact of small-scaleactions on environmental changes, and it also can solve the problem of identifying actions with thesame trajectory and different sizes by virtue of its multi-scale characteristics. Finally, we collected6240 instances for 52 kinds of handwritten letters from 6 volunteers. There are 3120 instances fromthe lab and 3120 instances are from the utility room. Using 10-fold cross-validation, the accuracyof MCSM-Wri is 95.31%, 96.68%, and 97.70% for the lab, the utility room, and the lab+utility room,respectively. Compared with Wi-Wri and SignFi, we increased the accuracy from 8.96% to 18.13% forrecognizing handwritten letters.

An Efficient Retinal Vessels Biometric Recognition System by Using Multi-Scale Local Binary Pattern Descriptor

2020 ◽  
Vol 10 (10) ◽  
pp. 2481-2489
Author(s):  
Muhammad Sheraz Arshad Malik ◽  
Qoseen Zahra ◽  
Imran Ullah Khan ◽  
Muhammad Awais ◽  
Gang Qiao
Keyword(s):  
Feature Extraction ◽  
Random Forest ◽  
Efficient Method ◽  
Large Scale ◽  
Local Binary Pattern ◽  
Evaluation Criteria ◽  
Recognition System ◽  
Suggested Approach ◽  
Multi Scale ◽  
Biometric Systems

Biometric systems are technically used for human recognition by identifying the unique features of an individual. Many security issues are found related to biometric systems such as voice, fingerprints, face, iris, signatures, etc., but the retina is a unique and efficient method to identify valid one. The aim of this paper is provided with an efficient method to recognize someone based on unique retina features. A proposed system based on retinal blood vessel pattern by using multi-scale local binary pattern (MSLBP) and random forest (Bagging tree) as feature extraction and classification. MSLBP is an efficient method to extracted features at six scales perpixel level, earlier work found the deficiency based on simple binary pattern with coverage of small areas and per-pixel level in the surrounding. MSLBP and random forest classifier suggested approach use for improving usability, perceivability, and sensitivity on large scale areas. It is the fastest method to get features accurately in an efficient way at every level of pixels. This method based on deep learning evaluation (criteria) parameter selection that provides more significant influence with sharp feature extraction on large scale areas based on seconds and improves the efficiency of images. MSLBP overcomes the problem of image sizing, pixel levels and efficiently provide accurate results.

APPLICATION OF COMPETITIVE HOPFIELD NEURAL NETWORK TO BRAIN-COMPUTER INTERFACE SYSTEMS

2012 ◽  
Vol 22 (01) ◽  
pp. 51-62 ◽  
Author(s):  
WEI-YEN HSU
Keyword(s):  
Neural Network ◽  
Hopfield Neural Network ◽  
Recognition System ◽  
Data Sets ◽  
Self Organizing Map ◽  
Time Frequency ◽  
Multi Scale ◽  
Eeg Data ◽  
Supervised Classifiers ◽  
Active Segment

We propose an unsupervised recognition system for single-trial classification of motor imagery (MI) electroencephalogram (EEG) data in this study. Competitive Hopfield neural network (CHNN) clustering is used for the discrimination of left and right MI EEG data posterior to selecting active segment and extracting fractal features in multi-scale. First, we use continuous wavelet transform (CWT) and Student's two-sample t-statistics to select the active segment in the time-frequency domain. The multiresolution fractal features are then extracted from wavelet data by means of modified fractal dimension. At last, CHNN clustering is adopted to recognize extracted features. Due to the characteristic of non-supervision, it is proper for CHNN to classify non-stationary EEG signals. The results indicate that CHNN achieves 81.9% in average classification accuracy in comparison with self-organizing map (SOM) and several popular supervised classifiers on six subjects from two data sets.

A Novel Finger-Vein Recognition Based on Quality Assessment and Multi-Scale Histogram of Oriented Gradients Feature

2019 ◽  
Vol 15 (1) ◽  
pp. 100-115 ◽  
Author(s):  
Junying Zeng ◽  
Yao Chen ◽  
Yikui Zhai ◽  
Junying Gan ◽  
Wulin Feng ◽  
...  
Keyword(s):  
Quality Assessment ◽  
Image Quality Assessment ◽  
Recognition System ◽  
Quality Score ◽  
Histogram Of Oriented Gradients ◽  
Test Results ◽  
Multi Scale ◽  
Finger Vein ◽  
Vein Recognition ◽  
Finger Vein Recognition

Inferior finger vein images would seriously alter the completion of recognition systems. A modern finger-vein recognition technique combined with image quality assessment is developed to overcome those drawbacks. By the quality assessment, this article can discard the inferior images and retain the superior images which are then transferred to the recognition system. Different from previous methods, this article assesses the quality features of the image for the purpose of distinguishing whether the image contains rich and stable vein characteristics. In light of this purpose, the quality assessment is implemented: first, the finger vein image is automatically annotated; second, the finger vein image is cut into image blocks to expand the training set; third, the average quality score of multiple image blocks from an image is the final quality score of the image in the course of testing. Next, the Histogram of Oriented Gradients (HOG) features are extracted from the four transformed high-quality sub-images, whose features are cascaded into the multi-scale HOG feature of an image. Finally, two modules, the quality assessment module using Convolutional Neural Networks (CNN) and finger vein recognition module which make full use of multi-scale HOG, are perfectly combined in this article. The test results have demonstrated that light-CNN can identifies inferior and superior images accurately and the multi-scale HOG is feasible and effective. What's more, this article can see the robustness of this combined method in this article.

scholarly journals Landmark Generation in Visual Place Recognition Using Multi-Scale Sliding Window for Robotics

2019 ◽  
Vol 9 (15) ◽  
pp. 3146 ◽  
Author(s):  
Bo Yang ◽  
Xiaosu Xu ◽  
Jun Li ◽  
Hong Zhang
Keyword(s):  
Recognition Performance ◽  
Input Parameter ◽  
Sliding Window ◽  
Recognition System ◽  
Place Recognition ◽  
Size Distributions ◽  
Multi Scale ◽  
Shift Invariance ◽  
Single Input ◽  
Visual Place Recognition

Landmark generation is an essential component in landmark-based visual place recognition. In this paper, we present a simple yet effective method, called multi-scale sliding window (MSW), for landmark generation in order to improve the performance of place recognition. In our method, we generate landmarks that form a uniform distribution in multiple landmark scales (sizes) within an appropriate range by a process that samples an image with a sliding window. This is in contrast to conventional methods of landmark generation that typically depend on detecting objects whose size distributions are uneven and, as a result, may not be effective in achieving shift invariance and viewpoint invariance, two important properties in visual place recognition. We conducted experiments on four challenging datasets to demonstrate that the recognition performance can be significantly improved by our method in a standard landmark-based visual place recognition system. Our method is simple with a single input parameter, the scales of landmarks required, and it is efficient as it does not involve detecting objects.

Sign in / Sign up

Export Citation Format

Share Document