scholarly journals Classification of Microscopic Laser Engraving Surface Defect Images Based on Transfer Learning Method

Electronics ◽  
2021 ◽  
Vol 10 (16) ◽  
pp. 1993
Author(s):  
Jing Zhang ◽  
Zhenhao Li ◽  
Ruqian Hao ◽  
Xiangzhou Wang ◽  
Xiaohui Du ◽  
...  
Keyword(s):  
Feature Extraction ◽  
Image Classification ◽  
Transfer Learning ◽  
Surface Defect ◽  
Feature Space ◽  
Fine Tuning ◽  
Defect Classification ◽  
Classification Methods ◽  
Laser Engraving ◽  
Traditional Classification

Microscopic laser engraving surface defect classification plays an important role in the industrial quality inspection field. The key challenges of accurate surface defect classification are the complete description of the defect and the correct distinction into categories in the feature space. Traditional classification methods focus on the terms of feature extraction and independent classification; therefore, feed handcrafted features may result in useful feature loss. In recent years, convolutional neural networks (CNNs) have achieved excellent results in image classification tasks with the development of deep learning. Deep convolutional networks integrate feature extraction and classification into self-learning, but require large datasets. The training datasets for microscopic laser engraving image classification are small; therefore, we used pre-trained CNN models and applied two fine-tuning strategies. Transfer learning proved to perform well even on small future datasets. The proposed method was evaluated on the datasets consisting of 1986 laser engraving images captured by a metallographic microscope and annotated by experienced staff. Because handcrafted features were not used, our method is more robust and achieves better results than traditional classification methods. Under five-fold-validation, the average accuracy of the best model based on DenseNet121 is 96.72%.

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.

scholarly journals Steel Surface Defect Classification Based on Small Sample Learning

2021 ◽  
Vol 11 (23) ◽  
pp. 11459
Author(s):  
Shiqing Wu ◽  
Shiyu Zhao ◽  
Qianqian Zhang ◽  
Long Chen ◽  
Chenrui Wu
Keyword(s):  
Feature Extraction ◽  
Steel Surface ◽  
Surface Defect ◽  
Nearest Neighbor ◽  
Surface Defects ◽  
Feature Space ◽  
Small Sample ◽  
Defect Classification ◽  
Sample Number ◽  
Degradation Problem

The classification of steel surface defects plays a very important role in analyzing their causes to improve manufacturing process and eliminate defects. However, defective samples are very scarce in actual production, so using very few samples to construct a good classifier is a challenge to be addressed. If the layer number of the model with proper depth is increased, the model accuracy will decrease (not caused by overfit), and the training error as well as the test error will be very high. This is called the degradation problem. In this paper, we propose to use feature extraction + feature transformation + nearest neighbors to classify steel surface defects. In order to solve the degradation problem caused by network deepening, the three feature extraction networks of Residual Net, Mobile Net and Dense Net are designed and analyzed. Experiment results show that in the case of a small sample number, Dense block can better solve the degradation problem caused by network deepening than Residual block. Moreover, if Dense Net is used as the feature extraction network, and the nearest neighbor classification algorithm based on Euclidean metric is used in the new feature space, the defect classification accuracy can reach 92.33% when only five labeled images of each category are used as the training set. This paper is of some guiding significance for surface defect classification when the sample number is small.

scholarly journals Aerial Scene Classification through Fine-Tuning with Adaptive Learning Rates and Label Smoothing

2020 ◽  
Vol 10 (17) ◽  
pp. 5792 ◽  
Author(s):  
Biserka Petrovska ◽  
Tatjana Atanasova-Pacemska ◽  
Roberto Corizzo ◽  
Paolo Mignone ◽  
Petre Lameski ◽  
...  
Keyword(s):  
Remote Sensing ◽  
Neural Networks ◽  
Feature Extraction ◽  
Image Classification ◽  
Transfer Learning ◽  
Remote Sensing Image ◽  
Learning Rate ◽  
Fine Tuning ◽  
Support Vector ◽  
Learning Rates

Remote Sensing (RS) image classification has recently attracted great attention for its application in different tasks, including environmental monitoring, battlefield surveillance, and geospatial object detection. The best practices for these tasks often involve transfer learning from pre-trained Convolutional Neural Networks (CNNs). A common approach in the literature is employing CNNs for feature extraction, and subsequently train classifiers exploiting such features. In this paper, we propose the adoption of transfer learning by fine-tuning pre-trained CNNs for end-to-end aerial image classification. Our approach performs feature extraction from the fine-tuned neural networks and remote sensing image classification with a Support Vector Machine (SVM) model with linear and Radial Basis Function (RBF) kernels. To tune the learning rate hyperparameter, we employ a linear decay learning rate scheduler as well as cyclical learning rates. Moreover, in order to mitigate the overfitting problem of pre-trained models, we apply label smoothing regularization. For the fine-tuning and feature extraction process, we adopt the Inception-v3 and Xception inception-based CNNs, as well the residual-based networks ResNet50 and DenseNet121. We present extensive experiments on two real-world remote sensing image datasets: AID and NWPU-RESISC45. The results show that the proposed method exhibits classification accuracy of up to 98%, outperforming other state-of-the-art methods.

scholarly journals A Multi-Domain Collaborative Transfer Learning Method with Multi-Scale Repeated Attention Mechanism for Underwater Side-Scan Sonar Image Classification

2022 ◽  
Vol 14 (2) ◽  
pp. 355
Author(s):  
Zhen Cheng ◽  
Guanying Huo ◽  
Haisen Li
Keyword(s):  
Feature Extraction ◽  
Deep Learning ◽  
Image Classification ◽  
Transfer Learning ◽  
Fine Tuning ◽  
Side Scan Sonar ◽  
Multi Scale ◽  
Sonar Image ◽  
Optical Images

Due to the strong speckle noise caused by the seabed reverberation which makes it difficult to extract discriminating and noiseless features of a target, recognition and classification of underwater targets using side-scan sonar (SSS) images is a big challenge. Moreover, unlike classification of optical images which can use a large dataset to train the classifier, classification of SSS images usually has to exploit a very small dataset for training, which may cause classifier overfitting. Compared with traditional feature extraction methods using descriptors—such as Haar, SIFT, and LBP—deep learning-based methods are more powerful in capturing discriminating features. After training on a large optical dataset, e.g., ImageNet, direct fine-tuning method brings improvement to the sonar image classification using a small-size SSS image dataset. However, due to the different statistical characteristics between optical images and sonar images, transfer learning methods—e.g., fine-tuning—lack cross-domain adaptability, and therefore cannot achieve very satisfactory results. In this paper, a multi-domain collaborative transfer learning (MDCTL) method with multi-scale repeated attention mechanism (MSRAM) is proposed for improving the accuracy of underwater sonar image classification. In the MDCTL method, low-level characteristic similarity between SSS images and synthetic aperture radar (SAR) images, and high-level representation similarity between SSS images and optical images are used together to enhance the feature extraction ability of the deep learning model. Using different characteristics of multi-domain data to efficiently capture useful features for the sonar image classification, MDCTL offers a new way for transfer learning. MSRAM is used to effectively combine multi-scale features to make the proposed model pay more attention to the shape details of the target excluding the noise. Experimental results of classification show that, in using multi-domain data sets, the proposed method is more stable with an overall accuracy of 99.21%, bringing an improvement of 4.54% compared with the fine-tuned VGG19. Results given by diverse visualization methods also demonstrate that the method is more powerful in feature representation by using the MDCTL and MSRAM.

scholarly journals Natural Images Allow Universal Adversarial Attacks on Medical Image Classification Using Deep Neural Networks with Transfer Learning

2021 ◽  
Author(s):  
Akinori Minagi ◽  
Hokuto Hirano ◽  
Kazuhiro Takemoto
Keyword(s):  
Neural Networks ◽  
Image Classification ◽  
Transfer Learning ◽  
Medical Image ◽  
Deep Neural Networks ◽  
Disease Diagnosis ◽  
Natural Images ◽  
Fine Tuning ◽  
Security And Privacy ◽  
Medical Image Classification

Abstract Transfer learning from natural images is well used in deep neural networks (DNNs) for medical image classification to achieve computer-aided clinical diagnosis. Although the adversarial vulnerability of DNNs hinders practical applications owing to the high stakes of diagnosis, adversarial attacks are expected to be limited because training data — which are often required for adversarial attacks — are generally unavailable in terms of security and privacy preservation. Nevertheless, we hypothesized that adversarial attacks are also possible using natural images because pre-trained models do not change significantly after fine-tuning. We focused on three representative DNN-based medical image classification tasks (i.e., skin cancer, referable diabetic retinopathy, and pneumonia classifications) and investigated whether medical DNN models with transfer learning are vulnerable to universal adversarial perturbations (UAPs), generated using natural images. UAPs from natural images are useful for both non-targeted and targeted attacks. The performance of UAPs from natural images was significantly higher than that of random controls, although slightly lower than that of UAPs from training images. Vulnerability to UAPs from natural images was observed between different natural image datasets and between different model architectures. The use of transfer learning causes a security hole, which decreases the reliability and safety of computer-based disease diagnosis. Model training from random initialization (without transfer learning) reduced the performance of UAPs from natural images; however, it did not completely avoid vulnerability to UAPs. The vulnerability of UAPs from natural images will become a remarkable security threat.

scholarly journals Improvement for Convolutional Neural Networks in Image Classification Using Long Skip Connection

2021 ◽  
Vol 11 (5) ◽  
pp. 2092
Author(s):  
Hong Hai Hoang ◽  
Hoang Hieu Trinh
Keyword(s):  
Neural Networks ◽  
Image Classification ◽  
Convolutional Neural Networks ◽  
Surface Defect ◽  
Spatial Knowledge ◽  
Floating Point ◽  
Defect Classification

In this paper, we examine and research the effect of long skip connection on convolutional neural networks (CNNs) for the tasks of image (surface defect) classification. The standard popular models only apply short skip connection inside blocks (layers with the same size). We apply the long version of residual connection on several proposed models, which aims to reuse the lost spatial knowledge from the layers close to input. For some models, Depthwise Separable Convolution is used rather than traditional convolution in order to reduce both count of parameters and floating-point operations per second (FLOPs). Comparative experiments of the newly upgraded models and some popular models have been carried out on different datasets including Bamboo strips datasets and a reduced version of ImageNet. The modified version of DenseNet 121 (we call MDenseNet 121) achieves higher validation accuracy while it has about 75% of weights and FLOPs in comparison to the original DenseNet 121.

scholarly journals How Deeply to Fine-Tune a Convolutional Neural Network: A Case Study Using a Histopathology Dataset

2020 ◽  
Vol 10 (10) ◽  
pp. 3359 ◽  
Author(s):  
Ibrahem Kandel ◽  
Mauro Castelli
Keyword(s):  
Image Classification ◽  
Transfer Learning ◽  
Medical Image ◽  
Disease Diagnosis ◽  
Fine Tuning ◽  
Medical Image Classification ◽  
Fine Tune ◽  
Entire Network

Accurate classification of medical images is of great importance for correct disease diagnosis. The automation of medical image classification is of great necessity because it can provide a second opinion or even a better classification in case of a shortage of experienced medical staff. Convolutional neural networks (CNN) were introduced to improve the image classification domain by eliminating the need to manually select which features to use to classify images. Training CNN from scratch requires very large annotated datasets that are scarce in the medical field. Transfer learning of CNN weights from another large non-medical dataset can help overcome the problem of medical image scarcity. Transfer learning consists of fine-tuning CNN layers to suit the new dataset. The main questions when using transfer learning are how deeply to fine-tune the network and what difference in generalization that will make. In this paper, all of the experiments were done on two histopathology datasets using three state-of-the-art architectures to systematically study the effect of block-wise fine-tuning of CNN. Results show that fine-tuning the entire network is not always the best option; especially for shallow networks, alternatively fine-tuning the top blocks can save both time and computational power and produce more robust classifiers.

scholarly journals Genetic Programming based Feature Manipulation for Skin Cancer Image Classification

2021 ◽  
Author(s):  
◽  
~ Qurrat Ul Ain
Keyword(s):  
Feature Extraction ◽  
Feature Selection ◽  
Image Classification ◽  
State Of The Art ◽  
Classification Performance ◽  
Image Features ◽  
Feature Construction ◽  
Classification Methods ◽  
Wrapper Approach ◽  
High Level

<p>Skin image classification involves the development of computational methods for solving problems such as cancer detection in lesion images, and their use for biomedical research and clinical care. Such methods aim at extracting relevant information or knowledge from skin images that can significantly assist in the early detection of disease. Skin images are enormous, and come with various artifacts that hinder effective feature extraction leading to inaccurate classification. Feature selection and feature construction can significantly reduce the amount of data while improving classification performance by selecting prominent features and constructing high-level features. Existing approaches mostly rely on expert intervention and follow multiple stages for pre-processing, feature extraction, and classification, which decreases the reliability, and increases the computational complexity. Since good generalization accuracy is not always the primary objective, clinicians are also interested in analyzing specific features such as pigment network, streaks, and blobs responsible for developing the disease; interpretable methods are favored. In Evolutionary Computation, Genetic Programming (GP) can automatically evolve an interpretable model and address the curse of dimensionality (through feature selection and construction). GP has been successfully applied to many areas, but its potential for feature selection, feature construction, and classification in skin images has not been thoroughly investigated. The overall goal of this thesis is to develop a new GP approach to skin image classification by utilizing GP to evolve programs that are capable of automatically selecting prominent image features, constructing new high level features, interpreting useful image features which can help dermatologist to diagnose a type of cancer, and are robust to processing skin images captured from specialized instruments and standard cameras. This thesis focuses on utilizing a wide range of texture, color, frequency-based, local, and global image properties at the terminal nodes of GP to classify skin cancer images from multiple modalities effectively. This thesis develops new two-stage GP methods using embedded and wrapper feature selection and construction approaches to automatically generating a feature vector of selected and constructed features for classification. The results show that wrapper approach outperforms the embedded approach, the existing baseline GP and other machine learning methods, but the embedded approach is faster than the wrapper approach. This thesis develops a multi-tree GP based embedded feature selection approach for melanoma detection using domain specific and domain independent features. It explores suitable crossover and mutation operators to evolve GP classifiers effectively and further extends this approach using a weighted fitness function. The results show that these multi-tree approaches outperformed single tree GP and other classification methods. They identify that a specific feature extraction method extracts most suitable features for particular images taken from a specific optical instrument. This thesis develops the first GP method utilizing frequency-based wavelet features, where the wrapper based feature selection and construction methods automatically evolve useful constructed features to improve the classification performance. The results show the evidence of successful feature construction by significantly outperforming existing GP approaches, state-of-the-art CNN, and other classification methods. This thesis develops a GP approach to multiple feature construction for ensemble learning in classification. The results show that the ensemble method outperformed existing GP approaches, state-of-the-art skin image classification, and commonly used ensemble methods. Further analysis of the evolved constructed features identified important image features that can potentially help the dermatologist identify further medical procedures in real-world situations.</p>

Sign in / Sign up

Export Citation Format

Share Document