scholarly journals Evaluation of Fine Tuning and Feature Extraction methods in Biometric Periocular Recognition

2019 ◽  
Author(s):  
William Barcellos ◽  
Nicolas Hiroaki Shitara ◽  
Carolina Toledo Ferraz ◽  
Raissa Tavares Vieira Queiroga ◽  
Jose Hiroki Saito ◽  
...  
Keyword(s):  
Feature Extraction ◽  
Transfer Learning ◽  
Class Imbalance ◽  
Extraction Methods ◽  
Fine Tuning ◽  
Feature Extraction Method ◽  
Network Training ◽  
Learning Techniques ◽  
Lower Accuracy ◽  
Different Characteristics

The aim of this paper is to evaluate the performance of Transfer Learning techniques applied in Convolucional Neural Networks for biometric periocular classification. Two aspects of Transfer Learning were evaluated: the technique known as Fine Tuning and the technique known as Feature Extraction. Two CNN architectures were evaluated, the AlexNet and the VGG-16, and two image databases were used. These two databases have different characteristics regarding the method of acquisition, the amount of classes, the class balancing, and the number of elements in each class. Three experiments were conducted to evaluate the performance of the CNNs. In the first experiment we measured the Feature Extraction accuracy, and in the second one we evaluated the Fine Tuning performance. In the third experiment, we used the AlexNet for Fine Tuning in one database, and then, the FC7 layer of this trained CNN was used for Feature Extraction in the other database. We concluded that the data quality (the presence or not of class samples in the training set), class imbalance (different number of elements in each class) and the selection method of the training and testing, directly influence the CNN accuracy. The Feature Extraction method, by being more simple and does not require network training, has lower accuracy than Fine Tuning. Furthermore, Fine Tuning a CNN with periocular's images from one database, doesn't increase the accuracy of this CNN in Feature Extraction mode for another periocular's database. The accuracy is quite similar to that obtained by the original pre-trained network

scholarly journals A Framework for Pedestrian Attribute Recognition Using Deep Learning

2022 ◽  
Vol 12 (2) ◽  
pp. 622
Author(s):  
Saadman Sakib ◽  
Kaushik Deb ◽  
Pranab Kumar Dhar ◽  
Oh-Jin Kwon
Keyword(s):  
Deep Learning ◽  
Transfer Learning ◽  
Class Imbalance ◽  
Recognition Task ◽  
Fine Tuning ◽  
Class Imbalance Problem ◽  
Imbalance Problem ◽  
Technological Advances ◽  
Learning Techniques ◽  
Attribute Recognition

The pedestrian attribute recognition task is becoming more popular daily because of its significant role in surveillance scenarios. As the technological advances are significantly more than before, deep learning came to the surface of computer vision. Previous works applied deep learning in different ways to recognize pedestrian attributes. The results are satisfactory, but still, there is some scope for improvement. The transfer learning technique is becoming more popular for its extraordinary performance in reducing computation cost and scarcity of data in any task. This paper proposes a framework that can work in surveillance scenarios to recognize pedestrian attributes. The mask R-CNN object detector extracts the pedestrians. Additionally, we applied transfer learning techniques on different CNN architectures, i.e., Inception ResNet v2, Xception, ResNet 101 v2, ResNet 152 v2. The main contribution of this paper is fine-tuning the ResNet 152 v2 architecture, which is performed by freezing layers, last 4, 8, 12, 14, 20, none, and all. Moreover, data balancing techniques are applied, i.e., oversampling, to resolve the class imbalance problem of the dataset and analysis of the usefulness of this technique is discussed in this paper. Our proposed framework outperforms state-of-the-art methods, and it provides 93.41% mA and 89.24% mA on the RAP v2 and PARSE100K datasets, respectively.

scholarly journals Comparison of different feature extraction methods for applicable automated ICD coding

2022 ◽  
Vol 22 (1) ◽  
Author(s):  
Zhao Shuai ◽  
Diao Xiaolin ◽  
Yuan Jing ◽  
Huo Yanni ◽  
Cui Meng ◽  
...  
Keyword(s):  
Machine Learning ◽  
Feature Extraction ◽  
Extraction Methods ◽  
Fine Tuning ◽  
Medical Texts ◽  
Optimal Method ◽  
Feature Extraction Method ◽  
Icd Codes ◽  
Shed Light ◽  
Icd Coding

Abstract Background Automated ICD coding on medical texts via machine learning has been a hot topic. Related studies from medical field heavily relies on conventional bag-of-words (BoW) as the feature extraction method, and do not commonly use more complicated methods, such as word2vec (W2V) and large pretrained models like BERT. This study aimed at uncovering the most effective feature extraction methods for coding models by comparing BoW, W2V and BERT variants. Methods We experimented with a Chinese dataset from Fuwai Hospital, which contains 6947 records and 1532 unique ICD codes, and a public Spanish dataset, which contains 1000 records and 2557 unique ICD codes. We designed coding tasks with different code frequency thresholds (denoted as $$f_s$$ f s ), with a lower threshold indicating a more complex task. Using traditional classifiers, we compared BoW, W2V and BERT variants on accomplishing these coding tasks. Results When $$f_s$$ f s was equal to or greater than 140 for Fuwai dataset, and 60 for the Spanish dataset, the BERT variants with the whole network fine-tuned was the best method, leading to a Micro-F1 of 93.9% for Fuwai data when $$f_s=200$$ f s = 200 , and a Micro-F1 of 85.41% for the Spanish dataset when $$f_s=180$$ f s = 180 . When $$f_s$$ f s fell below 140 for Fuwai dataset, and 60 for the Spanish dataset, BoW turned out to be the best, leading to a Micro-F1 of 83% for Fuwai dataset when $$f_s=20$$ f s = 20 , and a Micro-F1 of 39.1% for the Spanish dataset when $$f_s=20$$ f s = 20 . Our experiments also showed that both the BERT variants and BoW possessed good interpretability, which is important for medical applications of coding models. Conclusions This study shed light on building promising machine learning models for automated ICD coding by revealing the most effective feature extraction methods. Concretely, our results indicated that fine-tuning the whole network of the BERT variants was the optimal method for tasks covering only frequent codes, especially codes that represented unspecified diseases, while BoW was the best for tasks involving both frequent and infrequent codes. The frequency threshold where the best-performing method varied differed between different datasets due to factors like language and codeset.

scholarly journals Deep Learning Techniques for Grape Plant Species Identification in Natural Images

Sensors ◽  
2019 ◽  
Vol 19 (22) ◽  
pp. 4850 ◽  
Author(s):  
Carlos S. Pereira ◽  
Raul Morais ◽  
Manuel J. C. S. Reis
Keyword(s):  
Transfer Learning ◽  
Climatic Conditions ◽  
Fine Tuning ◽  
Variety Identification ◽  
Test Accuracy ◽  
Accuracy Score ◽  
Learning Techniques ◽  
Four Corners ◽  
Integrated Software ◽  
Grape Varieties

Frequently, the vineyards in the Douro Region present multiple grape varieties per parcel and even per row. An automatic algorithm for grape variety identification as an integrated software component was proposed that can be applied, for example, to a robotic harvesting system. However, some issues and constraints in its development were highlighted, namely, the images captured in natural environment, low volume of images, high similarity of the images among different grape varieties, leaf senescence, and significant changes on the grapevine leaf and bunch images in the harvest seasons, mainly due to adverse climatic conditions, diseases, and the presence of pesticides. In this paper, the performance of the transfer learning and fine-tuning techniques based on AlexNet architecture were evaluated when applied to the identification of grape varieties. Two natural vineyard image datasets were captured in different geographical locations and harvest seasons. To generate different datasets for training and classification, some image processing methods, including a proposed four-corners-in-one image warping algorithm, were used. The experimental results, obtained from the application of an AlexNet-based transfer learning scheme and trained on the image dataset pre-processed through the four-corners-in-one method, achieved a test accuracy score of 77.30%. Applying this classifier model, an accuracy of 89.75% on the popular Flavia leaf dataset was reached. The results obtained by the proposed approach are promising and encouraging in helping Douro wine growers in the automatic task of identifying grape varieties.

A Structural Analysis Based Feature Extraction Method for OCR System For Myanmar Printed Document Images

2012 ◽  
Vol 2 (1) ◽  
pp. 16-41 ◽  
Author(s):  
Htwe Pa Pa Win ◽  
Phyo Thu Thu Khine ◽  
Khin Nwe Ni Tun
Keyword(s):  
Feature Extraction ◽  
Structural Analysis ◽  
Character Recognition ◽  
Optical Character Recognition ◽  
Extraction Method ◽  
Recognition Performance ◽  
Extraction Methods ◽  
Support Vector ◽  
Svm Classifier ◽  
Feature Extraction Method

This paper proposes a new feature extraction method for off-line recognition of Myanmar printed documents. One of the most important factors to achieve high recognition performance in Optical Character Recognition (OCR) system is the selection of the feature extraction methods. Different types of existing OCR systems used various feature extraction methods because of the diversity of the scripts’ natures. One major contribution of the work in this paper is the design of logically rigorous coding based features. To show the effectiveness of the proposed method, this paper assumed the documents are successfully segmented into characters and extracted features from these isolated Myanmar characters. These features are extracted using structural analysis of the Myanmar scripts. The experimental results have been carried out using the Support Vector Machine (SVM) classifier and compare the pervious proposed feature extraction method.

Research on Feature Extraction Method Based on Rub-Impact Acoustic Emission Signals of Mechanical Seal End Faces

2010 ◽  
Vol 36 ◽  
pp. 68-74
Author(s):  
Chuan Jun Liao ◽  
Shuang Fu Suo ◽  
Wei Feng Huang
Keyword(s):  
Acoustic Emission ◽  
Feature Extraction ◽  
Extraction Methods ◽  
Mechanical Seal ◽  
Mechanical Seals ◽  
Feature Extraction Method ◽  
Different Types ◽  
Ae Signal ◽  
Feature Information ◽  
Ae Signals

Acoustic emission (AE) techniques are put forward to monitor rub-impacts between rotating rings and stationary rings of mechanical seals by this paper. By analyzing feature extraction methods of the typical rub-impact AE signal, the method combining of wavelet scalogram and power spectrum is found useful, and can used to attribute the feature information implicated in rub-impact AE signals of mechanical seal end faces. Both simulations and experimental research prove that the method is effective, and are used successfully to identify the typical features of different types of rub-impacts of mechanical seal end faces.

scholarly journals Dense-Res Net for Endoscopic Image Classification

2021 ◽  
Author(s):  
Quoc-Huy Trinh ◽  
Minh-Van Nguyen
Keyword(s):  
Feature Extraction ◽  
Image Classification ◽  
Transfer Learning ◽  
Fine Tuning ◽  
Anatomical Landmarks ◽  
Pathological Findings ◽  
Gi Tract ◽  
Target Class ◽  
Endoscopic Procedures ◽  
Endoscopic Image

We propose a method that configures Fine-tuning to a combination of backbone DenseNet and ResNet to classify eight classes showing anatomical landmarks, pathological findings, to endoscopic procedures in the GI tract. Our Technique depends on Transfer Learning which combines two backbones, DenseNet 121 and ResNet 101, to improve the performance of Feature Extraction for classifying the target class. After experiment and evaluating our work, we get accuracy with an F1 score of approximately 0.93 while training 80000 and test 4000 images.

Lung Disease Classification by Novel Shape-Based Feature Extraction and New Hybrid Genetic Approach

2016 ◽  
pp. 321-346
Author(s):  
Bhuvaneswari Chandran ◽  
P. Aruna ◽  
D. Loganathan
Keyword(s):  
Feature Extraction ◽  
Extraction Method ◽  
Lung Diseases ◽  
Hybrid Genetic Algorithm ◽  
Extraction Methods ◽  
Disease Classification ◽  
Filter Method ◽  
Svm Classifier ◽  
Genetic Approach ◽  
Feature Extraction Method

The purpose of the chapter is to present a novel method to classify lung diseases from the computed tomography images which assist physicians in the diagnosis of lung diseases. The method is based on a new approach which combines a proposed M2 feature extraction method and a novel hybrid genetic approach with different types of classifiers. The feature extraction methods performed in this work are moment invariants, proposed multiscale filter method and proposed M2 feature extraction method. The essential features which are the results of the feature extraction technique are selected by the novel hybrid genetic algorithm feature selection algorithms. Classification is performed by the support vector machine, multilayer perceptron neural network and Bayes Net classifiers. The result obtained proves that the proposed technique is an efficient and robust method. The performance of the proposed M2 feature extraction with proposed hybrid GA and SVM classifier combination achieves maximum classification accuracy.

scholarly journals Vibration Feature Analysis for Gas-Insulated Switchgear Mechanical Fault Detection under Varying Current

2020 ◽  
Vol 10 (3) ◽  
pp. 944 ◽  
Author(s):  
Ying Feng ◽  
Jianwen Wu
Keyword(s):  
Feature Extraction ◽  
Fault Detection ◽  
Extraction Methods ◽  
Mechanical Analysis ◽  
Linear Vibration ◽  
Feature Extraction Method ◽  
Gas Insulated Switchgear ◽  
Non Linear ◽  
Actual Operation ◽  
Mechanical Defect

As a key component to ensure the safe operation of the power grid, mechanical defect diagnosis technology of gas-insulated switchgear (GIS) during operation is often neglected. At present, GIS mechanical fault detection based on vibration information has not been developed. The main reason is that the excitation current is considerable but uncontrollable in the actual operation of GIS. It is difficult to eliminate the influence of excitation on the vibration amplitude and form an effective vibration feature description technology. Therefore, this paper proposes a unified feature-extraction method for GIS vibration information that reduces the influence of current amplitude for mechanical fault diagnosis. Starting from the GIS mechanical analysis, the periodicity of vibration excitation and the influence of amplitude are discussed. Then, combined with the non-linear characteristics of GIS systems and non-linear vibration theory, the multiplier frequency energy ratio (MFER) is designed to extract vibration-unified features of GIS for diagnosing the mechanical fault under different current levels. The diagnosis results of the experimental data with different feature-extraction methods show the applicability and superiority of the proposed method in the GIS’s mechanical fault-detection field based on vibration information.

scholarly journals FEATURE EXTRACTION BASED ON DIRECT CALCULATION OF MUTUAL INFORMATION

2007 ◽  
Vol 21 (07) ◽  
pp. 1213-1231 ◽  
Author(s):  
NOJUN KWAK
Keyword(s):  
Feature Extraction ◽  
Mutual Information ◽  
Direct Calculation ◽  
Extraction Methods ◽  
Descent Method ◽  
Gradient Descent Method ◽  
Probability Density Estimation ◽  
Classification Problems ◽  
Feature Extraction Method ◽  
Window Method

In many pattern recognition problems, it is desirable to reduce the number of input features by extracting important features related to the problems. By focusing on only the problem-relevant features, the dimension of features can be greatly reduced and thereby can result in a better generalization performance with less computational complexity. In this paper, we propose a feature extraction method for handling classification problems. The proposed algorithm is used to search for a set of linear combinations of the original features, whose mutual information with the output class can be maximized. The mutual information between the extracted features and the output class is calculated by using the probability density estimation based on the Parzen window method. A greedy algorithm using the gradient descent method is used to determine the new features. The computational load is proportional to the square of the number of samples. The proposed method was applied to several classification problems, which showed better or comparable performances than the conventional feature extraction methods.

scholarly journals How to Utilize My App Reviews? A Novel Topics Extraction Machine Learning Schema for Strategic Business Purposes

Entropy ◽  
2020 ◽  
Vol 22 (11) ◽  
pp. 1310
Author(s):  
Ioannis Triantafyllou ◽  
Ioannis C. Drivas ◽  
Georgios Giannakopoulos
Keyword(s):  
Machine Learning ◽  
Feature Extraction ◽  
Software Maintenance ◽  
Extraction Methods ◽  
Machine Learning Algorithms ◽  
Classification Problems ◽  
Case Scenario ◽  
Feature Extraction Method ◽  
Term Extraction ◽  
One Step

Acquiring knowledge about users’ opinion and what they say regarding specific features within an app, constitutes a solid steppingstone for understanding their needs and concerns. App review utilization helps project management teams to identify threads and opportunities for app software maintenance, optimization and strategic marketing purposes. Nevertheless, app user review classification for identifying valuable gems of information for app software improvement, is a complex and multidimensional issue. It requires foresight and multiple combinations of sophisticated text pre-processing, feature extraction and machine learning methods to efficiently classify app reviews into specific topics. Against this backdrop, we propose a novel feature engineering classification schema that is capable to identify more efficiently and earlier terms-words within reviews that could be classified into specific topics. For this reason, we present a novel feature extraction method, the DEVMAX.DF combined with different machine learning algorithms to propose a solution in app review classification problems. One step further, a simulation of a real case scenario takes place to validate the effectiveness of the proposed classification schema into different apps. After multiple experiments, results indicate that the proposed schema outperforms other term extraction methods such as TF.IDF and χ2 to classify app reviews into topics. To this end, the paper contributes to the knowledge expansion of research and practitioners with the purpose to reinforce their decision-making process within the realm of app reviews utilization.

Sign in / Sign up

Export Citation Format

Share Document