Efficient fusion of handcrafted and pre-trained CNNs features to classify melanoma skin cancer

Filali, Youssef; Khoukhi, Hasnae El; Sabri, My Abdelouahed; Aarab, Abdellah

doi:10.1007/s11042-020-09637-4

Cited by 49 publications

(21 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While other reviewed skin lesion classification methods were based on Optimized neutrosophic k-means (ONKM) in [74] Genetic algorithm for optimizing the value of α in α-mean operation in the neutrosophic set. Also, adaptive k-means and Random Fores in [76], Multiclass SVM in [77,78,91], SVM, ANN, KNN, and TDV in [83]. A buzzard optimization function extraction algorithm and SVM classifier provides accuracy 94.3% and the buzzard optimization for feature extraction is awesome.…”

Section: Discussionmentioning

confidence: 99%

“…Filali et al [91] suggested a potent lesional skin classification technique based on a mixture of the most potent DL architectures (VggNet, ResNet, GoogLeNet, and AlexNet) and features of handcrafted (color, texture, skeleton, and shape) in order to diagnostic melanoma cancer. The model extracts from each pre-trained CNN 1000 features, and all the pre-trained models extracted 4000 features.…”

Section: Fig 9 Densenet Architecturementioning

confidence: 99%

See 1 more Smart Citation

Skin Lesions Classification Using Deep Learning Techniques: Review

Kareem

Abdulazeez

Zeebaree

2021

AJRCoS

View full text Add to dashboard Cite

Skin cancer is a significant health problem. More than 123,000 new cases per year are recorded. Melanoma is the most popular type of skin cancer, leading to more than 9000 deaths annually in the USA. Skin disease diagnosis is getting difficult due to visual similarities. While Melanoma is the most common form of skin cancer, other pathology types are also fatal. Automatic melanoma screening systems will be useful in identifying those skin cancers more appropriately. Advances in technology and growth in computational capabilities have allowed machine learning and deep learning algorithms to analyze skin lesion images. Deep Convolutional Neural Networks (DCNNs) have achieved more encouraging results, yet faster systems for diagnosing fatal diseases are the need of the hour. This paper presents a survey of techniques for skin cancer detection from images. The paper aims to present a review of existing state-of-the-art and effective models for automatically detecting Melanoma from skin images. The result of classifications and segmentation from the skin lesion images will be processed better using the ensemble deep learning algorithm.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Fig 9 Densenet Architecturementioning

confidence: 99%

Skin Lesions Classification Using Deep Learning Techniques: Review

Kareem

Abdulazeez

Zeebaree

2021

AJRCoS

View full text Add to dashboard Cite

show abstract

“…Although it has already been published that the effectiveness of DNNs outperforms those of hand-crafted descriptors, these methods were shown to be better at discriminating stationary textures under steady imaging conditions and proved more robust than DNN-based features to, for example, image rotation [ 61 ]. Moreover, the concatenation of handcrafted features (shape, skeleton, color, and texture) and features extracted from the most powerful deep learning architectures followed by classification using for standard classifiers (e.g., support vector machines) was shown to offer high classification performance [ 62 ]. Future efforts would be directed to confirm if our approach offers any advantages compared to the hand-designed methods.…”

Section: Discussionmentioning

confidence: 99%

Deep Neural Frameworks Improve the Accuracy of General Practitioners in the Classification of Pigmented Skin Lesions

Lucius

All

et al. 2020

Diagnostics

View full text Add to dashboard Cite

This study evaluated whether deep learning frameworks trained in large datasets can help non-dermatologist physicians improve their accuracy in categorizing the seven most common pigmented skin lesions. Open-source skin images were downloaded from the International Skin Imaging Collaboration (ISIC) archive. Different deep neural networks (DNNs) (n = 8) were trained based on a random dataset constituted of 8015 images. A test set of 2003 images was used to assess the classifiers’ performance at low (300 × 224 RGB) and high (600 × 450 RGB) image resolution and aggregated data (age, sex and lesion localization). We also organized two different contests to compare the DNN performance to that of general practitioners by means of unassisted image observation. Both at low and high image resolution, the DNN framework differentiated dermatological images with appreciable performance. In all cases, the accuracy was improved when adding clinical data to the framework. Finally, the least accurate DNN outperformed general practitioners. The physician’s accuracy was statistically improved when allowed to use the output of this algorithmic framework as guidance. DNNs are proven to be high performers as skin lesion classifiers and can improve general practitioner diagnosis accuracy in a routine clinical scenario.

show abstract

“…Feature fusion has been widely adopted by researchers for detection and classification tasks relevant to computed vision [45]- [49]. We have used GIST descriptors with default value.…”

Section: ) Local Binary Pattern (Lbp)mentioning

confidence: 99%

Classification and Analysis of Android Malware Images Using Feature Fusion Technique

et al. 2021

View full text Add to dashboard Cite

The super packed functionalities and artificial intelligence (AI)-powered applications have made the Android operating system a big player in the market. Android smartphones have become an integral part of life and users are reliant on their smart devices for making calls, sending text messages, navigation, games, and financial transactions to name a few. This evolution of the smartphone community has opened new horizons for malware developers. As malware variants are growing at a tremendous rate every year, there is an urgent need to combat against stealth malware techniques. This paper proposes a visualization and machine learning-based framework for classifying Android malware. Android malware applications from the DREBIN dataset were converted into grayscale images. In the first phase of the experiment, the proposed framework transforms Android malware into fifteen different image sections and identifies malware files by exploiting handcrafted features associated with Android malware images. The algorithms such as Gray Level Co-occurrence Matrix-based (GLCM), Global Image deScripTors (GIST), and Local Binary Pattern (LBP) are used to extract the handcrafted features from the image sections. The extracted features were further classified using machine learning algorithms like K-Nearest Neighbors, Support Vector Machines, and Random Forests. In the second phase of the experiment, handcrafted features were fused with CNN features to form the feature fusion strategy. The classification performance was evaluated against every malware image file section. The results obtained using the Feature Fusion strategy are compared with handcrafted features results. The experiment results conclude to the fact that Feature Fusion-SVM model is most suited for the identification and classification of Android malware using the certificate and Android Manifest (CR+AM) malware images. It attained an high accuracy of 93.24%.

show abstract

Efficient fusion of handcrafted and pre-trained CNNs features to classify melanoma skin cancer

Cited by 49 publications

References 51 publications

Skin Lesions Classification Using Deep Learning Techniques: Review

Skin Lesions Classification Using Deep Learning Techniques: Review

Deep Neural Frameworks Improve the Accuracy of General Practitioners in the Classification of Pigmented Skin Lesions

Classification and Analysis of Android Malware Images Using Feature Fusion Technique

Contact Info

Product

Resources

About