scholarly journals Shape and Boundary Similarity Features for Accurate HCC Image Recognition

2017 ◽  
Vol 2017 ◽  
pp. 1-12
Author(s):  
Xiaoyu Duan ◽  
Huiyan Jiang ◽  
Siqi Li
Keyword(s):  
Cross Validation ◽  
Classification Model ◽  
Feature Points ◽  
Nucleus Shape ◽  
Comparison Results ◽  
Manual Selection ◽  
Validation Tests ◽  
Information Difference ◽  
Optimal Feature ◽  
Selection Of

Nucleus morphology is of great importance in conventional cancer pathological diagnosis, which could provide information difference between normal and abnormal nuclei visually. Therefore, this paper proposes two novel kinds of features for normal and hepatocellular carcinoma (HCC) nucleus recognition, including shape and boundary similarity. First, each individual nucleus patch with the fixed size is obtained using center-proliferation segmentation (CPS) method. Then, nucleus shape library is constructed based on manual selection by pathologists, which is utilized to measure nucleus shape similarity via Dice, Jaccard, precision, and recall coefficients. Meanwhile, boundary similarity is evaluated through triangles composed of some boundary feature points for each nucleus. Finally, the conventional random forest (RF) is used to train and test the classification model for HCC nucleus recognition. Extensive cross-validation tests could facilitate the selection of the optimal feature set and the experiment comparison results demonstrate that our proposed morphological features are more beneficial for classification compared with other traditional characteristics.

Study about Recognition of Digital Meter Dial Reading Based on SVM

2014 ◽  
Vol 615 ◽  
pp. 194-197
Author(s):  
Zhen Yuan Tu ◽  
Fang Hua Ning ◽  
Wu Jia Yu
Keyword(s):  
Support Vector Machine ◽  
Cross Validation ◽  
Recognition Rate ◽  
Classification Model ◽  
Support Vector ◽  
Feature Points ◽  
Svm Classification ◽  
Performance Factors ◽  
Model Combining ◽  
Fast Recognition

In practice, it is difficult for Support Vector Machine (SVM) to have a relatively high recognition rate as well as a quite fast recognition speed. In order to resolve this defect, in this paper we build a SVM classification model combining numerical characteristics. We use readings of rotary natural meters as the test temple, do positioning, preprocessing, feature points extracting, classifying and other series of operations to the numeric region of the dial. Then with the idea of cross-validation, we keep doing parameter optimation to SVM. At last, after making a comprehensive contrast of the effects which numerous performance factors make on the experimental outputs, we try to give our explanation of the outputs from different perspectives.

An Improved Three-Stage Classifier for Activity Recognition

2017 ◽  
Vol 32 (01) ◽  
pp. 1860003 ◽  
Author(s):  
Enrique Garcia-Ceja ◽  
Ramon F. Brena
Keyword(s):  
Activity Recognition ◽  
Cross Validation ◽  
Inertial Sensors ◽  
Wearable Sensors ◽  
Research Area ◽  
Classification Model ◽  
Homogeneous Groups ◽  
Wide Range ◽  
Important Research Area ◽  
Selection Of

Recently, Human Activity Recognition (HAR) has become an important research area because of its wide range of applications in several domains such as health care, elder care, sports monitoring systems, etc. The use of wearable sensors — specifically the use of inertial sensors such as accelerometers and gyroscopes — has become the most common approach to recognize physical activities because of their unobtrusiveness and ubiquity. Overall, the process of building a HAR system starts with a feature extraction phase and then a classification model is trained. In the work of Siirtola et al. is proposed an intermediate clustering step to find the homogeneous groups of activities. For the recognition step, an instance is assigned to one of the groups and the final classification is performed inside that group. In this work we evaluate the clustering-based approach for activity classification proposed by Siirtola with two additional improvements: automatic selection of the number of groups and an instance reassignment procedure. In the original work, they evaluated their method using decision trees on a sports activities dataset. For our experiments, we evaluated seven different classification models on four public activity recognition datasets. Our results with 10-fold Cross Validation showed that the method proposed by Siirtola with our additional two improvements performed better in the majority of cases as compared to using the single classification model under consideration. When using Leave One User Out Cross Validation (user independent model) we found no differences between the proposed method and the single classification model.

scholarly journals IMPROVING PERFORMANCE OF FEATURE EXTRACTION IN SFM ALGORITHMS FOR 3D SPARSE POINT CLOUD

2019 ◽  
Vol XLII-2/W17 ◽  
pp. 101-106
Author(s):  
F. Condorelli ◽  
R. Higuchi ◽  
S. Nasu ◽  
F. Rinaudo ◽  
H. Sugawara
Keyword(s):  
Point Cloud ◽  
Point Clouds ◽  
Low Density ◽  
Feature Points ◽  
Historical Archives ◽  
Before And After ◽  
Manual Selection ◽  
Art And Architecture ◽  
Selection Of

Abstract. The use of Structure-from-Motion algorithms is a common practice to obtain a rapid photogrammetric reconstruction. However, the performance of these algorithms is limited by the fact that in some conditions the resulting point clouds present low density. This is the case when processing materials from historical archives, such as photographs and videos, which generates only sparse point clouds due to the lack of necessary information in the photogrammetric reconstruction. This paper explores ways to improve the performance of open source SfM algorithms in order to guarantee the presence of strategic feature points in the resulting point cloud, even if sparse. To reach this objective, a photogrammetric workflow is proposed to process historical images. The first part of the workflow presents a method that allows the manual selection of feature points during the photogrammetric process. The second part evaluates the metric quality of the reconstruction on the basis of a comparison with a point cloud that has a different density from the sparse point cloud. The workflow was applied to two different case studies. Transformations of wall paintings of the Karanlık church in Cappadocia were analysed thanks to the comparison of 3D model resulting from archive photographs and a recent survey. Then a comparison was performed between the state of the Komise building in Japan, before and after restoration. The findings show that the method applied allows the metric scale and evaluation of the model also in bad condition and when only low-density point clouds are available. Moreover, this tool should be of great use for both art and architecture historians and geomatics experts, to study the evolution of Cultural Heritage.

scholarly journals A Spectral Pitch Class Model of the Probe Tone Data and Scalic Tonality

2015 ◽  
Vol 32 (4) ◽  
pp. 364-393 ◽  
Author(s):  
Andrew J. Milne ◽  
Robin Laney ◽  
David B. Sharp
Keyword(s):  
Empirical Data ◽  
Cross Validation ◽  
Musical Culture ◽  
Class Model ◽  
Probe Tone ◽  
Pitch Class ◽  
Form Part ◽  
Validation Tests ◽  
Selection Of ◽  
Tonic Triad

In this paper, we introduce a small family of novel bottom-up (sensory) models of the Krumhansl and Kessler (1982) probe tone data. The models are based on the spectral pitch class similarities between all twelve pitch classes and the tonic degree and tonic triad. Cross-validation tests of a wide selection of models show ours to have amongst the highest fits to the data. We then extend one of our models to predict the tonics of a variety of different scales such as the harmonic minor, melodic minor, and harmonic major. The model produces sensible predictions for these scales. Furthermore, we also predict the tonics of a small selection of microtonal scales—scales that do not form part of any musical culture. These latter predictions may be tested when suitable empirical data have been collected.

Energy Efficiency Analysis in ERA using NFL-BA Algorithm

2020 ◽  
pp. 340-348
Author(s):  
Palky Mehta ◽  
H. L. Sharma
Keyword(s):  
Cluster Head ◽  
Vital Role ◽  
Wireless Sensor ◽  
Energy Usage ◽  
Cluster Heads ◽  
Comparison Results ◽  
Current Scenario ◽  
Energy Efficiency Analysis ◽  
Selection Of

In the current scenario of Wireless Sensor Network (WSN), power consumption is the major issue associated with nodes in WSN. LEACH technique plays a vital role of clustering in WSN and reduces the energy usage effectively. But LEACH has its own limitation in order to search cluster head nodes which are randomly distributed over the network. In this paper, ERA-NFL- BA algorithm is being proposed for selects the cluster heads in WSN. This algorithm help in selection of cluster heads can freely transform from global search to local search. At the end, a comparison has been done with earlier researcher using protocol ERA-NFL, which clearly shown that proposed Algorithm is best suited and from comparison results that ERA-NFL-BA has given better performance.

Safety Evaluation of Destination Lighting Treatment at Stop Controlled Cross-Intersections

2018 ◽  
Vol 2672 (16) ◽  
pp. 113-121 ◽  
Author(s):  
Amrita Goswamy ◽  
Shauna Hallmark ◽  
Theresa Litteral ◽  
Michael Pawlovich
Keyword(s):  
Rural Areas ◽  
Negative Binomial ◽  
Safety Evaluation ◽  
Negative Binomial Regression ◽  
State Agencies ◽  
Crash Frequency ◽  
Binomial Regression ◽  
Manual Selection ◽  
Roadway Lighting ◽  
Selection Of

Intersection crashes during nighttime hours may occur because of poor driver visual cognition of conflicting traffic or intersection presence. In rural areas, the only source of lighting is typically provided by vehicle headlights. Roadway lighting enhances driver recognition of intersection presence and visibility of signs and markings. Destination lighting provides some illumination for the intersection but is not intended to fully illuminate all approaches. Destination lighting has been widely used in Iowa but the effectiveness has not been well documented. This study, therefore, sought to evaluate the effect on safety of destination lighting at rural intersections. As part of an extensive data collection effort, locations with destination/street lighting were gathered with the assistance of several state agencies. After manual selection of a similar number of control intersections, propensity score matching using the caliper width technique was used to match 245 treatments with 245 control sites. Negative binomial regression was used to evaluate crash frequency data. The presence of destination lighting at stop-controlled cross-intersections generally reduced the night-to-day crash ratio by 19%. The presence of treatment or destination lighting was associated with a 33%–39% increase in daytime crashes across all models but was associated with an 18%–33% reduction in nighttime crashes. Injuries in nighttime crashes decreased by 24% and total nighttime crashes reduced by 33%. Property damage crashes were reduced by 18%.

scholarly journals Komparasi Metode Klasifikasi untuk Deteksi Ekspresi Wajah Dengan Fitur Facial Landmark

2021 ◽  
Vol 8 (5) ◽  
pp. 949
Author(s):  
Fitra A. Bachtiar ◽  
Muhammad Wafi
Keyword(s):  
Facial Expression ◽  
Facial Expressions ◽  
Cross Validation ◽  
Classification Model ◽  
Expression Recognition ◽  
Classification Methods ◽  
Human Machine Interaction ◽  
Facial Landmark ◽  
Total Data ◽  
Machine Interaction

<p><em>Human machine interaction</em>, khususnya pada <em>facial</em> <em>behavior</em> mulai banyak diperhatikan untuk dapat digunakan sebagai salah satu cara untuk personalisasi pengguna. Kombinasi ekstraksi fitur dengan metode klasifikasi dapat digunakan agar sebuah mesin dapat mengenali ekspresi wajah. Akan tetapi belum diketahui basis metode klasifikasi apa yang tepat untuk digunakan. Penelitian ini membandingkan tiga metode klasifikasi untuk melakukan klasifikasi ekspresi wajah. Dataset ekspresi wajah yang digunakan pada penelitian ini adalah JAFFE dataset dengan total 213 citra wajah yang menunjukkan 7 (tujuh) ekspresi wajah. Ekspresi wajah pada dataset tersebut yaitu <em>anger</em>, <em>disgust</em>, <em>fear</em>, <em>happy</em>, <em>neutral</em>, <em>sadness</em>, dan <em>surprised</em>. Facial Landmark digunakan sebagai ekstraksi fitur wajah. Model klasifikasi yang digunakan pada penelitian ini adalah ELM, SVM, dan <em>k</em>-NN. Masing masing model klasifikasi akan dicari nilai parameter terbaik dengan menggunakan 80% dari total data. 5- <em>fold</em> <em>cross-validation</em> digunakan untuk mencari parameter terbaik. Pengujian model dilakukan dengan 20% data dengan metode evaluasi akurasi, F1 Score, dan waktu komputasi. Nilai parameter terbaik pada ELM adalah menggunakan 40 hidden neuron, SVM dengan nilai  = 10<sup>5</sup> dan 200 iterasi, sedangkan untuk <em>k</em>-NN menggunakan 3 <em>k</em> tetangga. Hasil uji menggunakan parameter tersebut menunjukkan ELM merupakan algoritme terbaik diantara ketiga model klasifikasi tersebut. Akurasi dan F1 Score untuk klasifikasi ekspresi wajah untuk ELM mendapatkan nilai akurasi sebesar 0.76 dan F1 Score 0.76, sedangkan untuk waktu komputasi membutuhkan waktu 6.97´10<sup>-3</sup> detik.   </p><p> </p><p><em><strong>Abstract</strong></em></p><p class="Abstract">H<em>uman-machine interaction, especially facial behavior is considered to be use in user personalization. Feature extraction and classification model combinations can be used for a machine to understand the human facial expression. However, which classification base method should be used is not yet known. This study compares three classification methods for facial expression recognition. JAFFE dataset is used in this study with a total of 213 facial images which shows seven facial expressions. The seven facial expressions are anger, disgust, fear, happy, neutral, sadness, dan surprised. Facial Landmark is used as a facial component features. The classification model used in this study is ELM, SVM, and k-NN. The hyperparameter of each model is searched using 80% of the total data. 5-fold cross-validation is used to find the hyperparameter. The testing is done using 20% of the data and evaluated using accuracy, F1 Score, and computation time. The hyperparameter for ELM is 40 hidden neurons, SVM with  = 105 and 200 iteration, while k-NN used 3 k neighbors. The experiment results show that ELM outperforms other classification methods. The accuracy and F1 Score achieved by ELM is 0.76 and 0.76, respectively. Meanwhile, time computation takes 6.97 10<sup>-3</sup> seconds.      </em></p>

Sign in / Sign up

Export Citation Format

Share Document