Feature Combination Techniques Research Articles

Augmented Attention: Enhancing Morph Detection in Face Recognition

Face Morphing is a technique that involves blending two or more faces to create new often realistic- looking images. These morphed images are generated or created using morphing techniques or photo manipulation tools pose a significant peril to face recognition systems. In this paper, we proposed a method to improve deep learning based face morphing detection systems to be more robust against face morphing attacks. Leveraging deep learning models and algorithms including MTCNN (Multitask Cascaded Convolutional Neural Network) for efficient face extraction from original and morphed faces with a high accuracy and FaceNet for extracting unified embeddings from the faces. The project aims to push the boundaries of face morphing detection capabilities by leveraging feature combination techniques using cosine distance and SSIM(structural similarity index measure) for identifying the similarity between faces and applying spatial attention mechanism which aims to enhance the feature representation learned by the model by focusing on the most informative parts of the image and training support vector machine classifier and voting classifier using the extracted embeddings significantly helps in building a robust face morphing detection system.

Improving Recognition of SIBI Gesture by Combining Skeleton and Hand Shape Features

SIBI (Sign System for Indonesian Language) is an official sign language system used in school for hearing impairment students in Indonesia. This work uses the skeleton and hand shape features to classify SIBI gestures. In order to improve the performance of the gesture classification system, we tried to fuse the features in several different ways. The accuracy results achieved by the feature fusion methods are, in descending order of accuracy: 88.016%, when using sequence-feature-vector concatenation, 85.448% when using Conneau feature vector concatenation, 83.723% when using feature-vector concatenation, and 49.618% when using simple feature concatenation. The sequence-feature-vector concatenation techniques yield noticeably better results than those achieved using single features (82.849% with skeleton feature only, 55.530% for the hand shape feature only). The experiment results show that the combined features of the whole gesture sequence can better distinguish one gesture from another in SIBI than the combined features of each gesture frame. In addition to finding the best feature combination technique, this study also found the most suitable Recurrent Neural Network (RNN) model for recognizing SIBI. The models tested are 1-layer, 2-layer LSTM, and GRU. The experimental results show that the 2-layer bidirectional LSTM has the best performance.

A novel hybrid feature combination method for enhanced movie recommendations with user resemblance and attitude mining.

Most movie recommendation methods use hard-clustering and simple collaborative filtering techniques in order to achieve their end results. However, these methods tend to overlook crucial aspects of both users and items. When these methods hard cluster a movie item into a cluster, they turn a blind eye to the fact that the item also exhibits some properties of another cluster’s items. Recommender systems facilitate users and relevant things expeditiously supported their requests and historic communications with alternative customers. Recommendation systems are a crucial portion of signifying things particularly in streaming amenities. For streaming motion-picture show services like Netflix, recommendation methods are vital for serving to users notice fresh movies to get pleasure from. However, massive amounts of information will turn out restrictions in recommendations due to accuracy as a result of diversity and meagerness problems. Our work proposes a unique hybrid technique that mixes collaborative filtering and characteristics of demographic filtering technique to point the close users, and associate against one another. This technique has been established over associate in tending analysis of the way to cut back the blunders in grading estimates supported users’ earlier communications that ends up in improved prediction accuracy in among completely different algorithms. Additionally, a feature combination technique is utilized that progresses the expectation accuracy and to check our method, using MovieLens 1M dataset, we contended an offline assessment, already available assessment tactics, and compared the same with the output factors to support authenticating the proposed procedure.

Efficient feature combination techniques for emotional speech classification

The challenge to enhance the naturalness and efficiency of spoken language man---machine interface, emotional speech identification and its classification has been a predominant research area. The reliability and accuracy of such emotion identification greatly depends on the feature selection and extraction. In this paper, a combined feature selection technique has been proposed which uses the reduced features set artifact of vector quantizer (VQ) in a Radial Basis Function Neural Network (RBFNN) environment for classification. In the initial stage, Linear Prediction Coefficient (LPC) and time---frequency Hurst parameter (pH) are utilized to extract the relevant feature, both exhibiting complementary information from the emotional speech. Extensive simulations have been carried out using Berlin Database of Emotional Speech (EMO-DB) with various combination of feature set. The experimental results reveal 76 % accuracy for pH and 68 % for LPC using standalone feature set, whereas the combination of feature sets, (LP VQC and pH VQC) enhance the average accuracy level up to 90.55 %.

Personal Identification Using Ears Based on Statistical Features

Biometrics is an automated method of recognizing a person based on a physiological (e.g. face, iris, or retina) or behavioral (e.g. gait, signature, or dynamic keystrokes) characteristics. Ear recognition is one of the physiological biometrics' types that have been interested in the recent years. Ear recognition, achieves good accuracy and has many advantages such as it doesn't affected by expressions, health, and more stable than many other biometrics. However, it has many challenges such as the pose of the face, lighting variation, occlusion with hair or clothes. In this research, four proposed models are used to identify people using ear images. The first model used single feature extraction method based on single classifier. While, the second model used single feature extraction method based on multi-classifiers. The third model used feature combination techniques (parallel or serial) based on single classifier. Finally, in the fourth model multi-features and multi-classifiers are used. In this research, there are four methods that are used to extract the features, namely, \textit{Principal Component Analysis} (PCA), \textit{Linear Discriminant Analysis} (LDA), \textit{Independent Component Analysis} (ICA), and \textit{Discrete Cousin Transform} (DCT). Neural networks, decision tree, and minimum distance classifiers are used to classify the unknown samples. The occlusion problem with hair or scarves is one of the big challenges of the ear recognition systems. In this research, segmentation technique is proposed to neglect the occluded part and solve the occlusion problem. The idea of the segmentation technique is based on dividing the ear images into different parts. The occluded part/s is neglected and the rest of the parts are used to identify people based on features fusion and classifiers fusion. The segmentation technique consists of two main types, namely, uniform or non-uniform segmentation techniques. In this research, the uniform segmentation technique is used for many experiments (horizontal, vertical, and grid). All the four proposed models are applied to all ear segments to investigate the power of each model and to achieve a high accuracy. In this research, ear database images is used. The ear dataset consists of 102 grayscale images (6 images for each of 17 subjects) in PGM format [1]. The proposed models are achieved good identification rates using ear images. In the first model, the best accuracy achieved using LDA and neural network classifier. The results of the first model ranged from 64.12\% to 100\%. In the second model, many classifiers are fused to increase the recognition rate. In this method, two methods are used, namely, Borda count and majority voting. The results of this model ranged from 94.12\% to 96.08\%. The third model, the features using two different methods, namely serial and parallel are combined. The results of this model prove that the serial combination is more powerful than parallel combination. Finally, in the fourth model, two features and two classifiers are fused to get one decision. The accuracy of this model is approximately the same of the third model, and it does not achieve good results because there is a diversity between different classifiers. Moreover, the proposed segmentation model achieved good results when some parts of the ear images are occluded.

Towards cost reduction of breast cancer diagnosis using mammography texture analysis

In this paper we analyse the performance of various texture analysis methods for the purpose of reducing the number of false positives in breast cancer detection; as a result, the cost of breast cancer diagnosis would be reduced. We consider well-known methods such as local binary patterns, histogram of oriented gradients, co-occurrence matrix features and Gabor filters. Moreover, we propose the use of local directional number patterns as a new feature extraction method for breast mass detection. For each method, different classifiers are trained on the extracted features to predict the class of unknown instances. In order to improve the mass detection capability of each individual method, we use feature combination techniques and classifier majority voting. Some experiments were performed on the images obtained from a public breast cancer database, achieving promising levels of sensitivity and specificity.

Vineyard identification in an oak woodland landscape with airborne digital camera imagery

Using airborne multispectral digital camera imagery, we compared a number of feature combination techniques in image classification to distinguish vineyard from non-vineyard land-cover types in northern California. Image processing techniques were applied to raw images to generate feature images including grey level co-occurrence based texture measures, low pass and Laplacian filtering results, Gram-Schmidt orthogonalization, principal components, and normalized difference vegetation index (NDVI). We used the maximum likelihood classifier for image classification. Accuracy assessment is performed using digitized boundaries of the vineyard blocks. The most successful classification as determined by t-tests of the Kappa coefficients was achieved based on the use of a texture image of homogeneity obtained from the near infrared image band, NDVI and brightness generated through orthogonalization analysis. This method averaged an overall accuracy of 81 per cent for six frames of images tested. With post-classification morphological processing (clumping and sieving) the overall accuracy was significantly increased to 87 per cent (with a confidence level of 0.99).